Щоб переглянути інші типи публікацій з цієї теми, перейдіть за посиланням: Machine learning interactif.
Дисертації з теми "Machine learning interactif"
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся з топ-50 дисертацій для дослідження на тему "Machine learning interactif".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Переглядайте дисертації для різних дисциплін та оформлюйте правильно вашу бібліографію.
1
Scurto, Hugo. "Designing With Machine Learning for Interactive Music Dispositifs." Electronic Thesis or Diss., Sorbonne université, 2019. http://www.theses.fr/2019SORUS356.
Повний текст джерелаАнотація:
La musique est une pratique culturelle permettant aux êtres humains d'exprimer sensiblement leurs intentions à travers le son. L'apprentissage machine définit un ensemble de modèles permettant de nouvelles formes d'expression au sein desdits systèmes interactifs musicaux. Cependant, en tant que discipline informatique, l'apprentissage machine demeure essentiellement appliquée à la musique du point de vue des sciences de l'ingénieur, qui, très souvent, conçoit les modèles d'apprentissage sans tenir compte des interactions musicales prenant place entre humains et systèmes. Dans cette thèse, j'envisage la possibilité de mener des pratiques de design avec l'apprentissage machine pour les systèmes interactifs musicaux. Je soutiens que l'apprentissage machine doit avant tout être situé au sein d'un contexte humain afin d'être conçu et appliqué au design de systèmes interactifs musicaux. Pour défendre cette thèse, je présente quatre études interdisciplinaires, dans lesquelles j'introduis des modèles intermédiaires d'apprentissage, dits modèles-prototype, au sein de méthodes de conception centrées humain, afin d'appliquer l'apprentissage machine à quatre tâches musicales situées : le mapping mouvement-son, l'exploration sonore, l'exploration de la synthèse, et l'interaction musicale collective. À travers ces études, je montre que les modèles-prototype permettent de générer des idées de design pour l'apprentissage machine en amont de la phase d'ingénierie desdits modèles, ce en lien étroit avec les utilisateurs potentiels de ces systèmes. Je montre également que les systèmes d'apprentissage machine centrés humain résultant de ce processus de conception rendent possible des processus dynamiques d'expression entre les humains et les machines, allant au-delà de la création d'artefacts musicaux statiques. Je propose de nommer co-expression ces processus d'interaction musicale entre des êtres humains - faisant preuve d'un élan expressif et créatif quelque soit leur expertise musicale - et des machines - dont les capacités d'apprentissage peuvent être perçues comme expressives du point de vue de l'humain. En outre, je présente cinq systèmes interactifs musicaux conçus avec lesdits modèles-prototypes, et relate leurs restitutions publiques au sein d'ateliers, expositions, installations et performances. Par une approche réflexive, je montre que les contributions musicales apportées par des pratiques de design avec l'apprentissage machine peuvent, à terme, complémenter les contributions scientifiques apportées par les méthodes de conception centrées humain. Ainsi, je suggère que la recherche musicale peut être menée par le design de dispositifs interactifs musicaux, c'est-à-dire, par la réalisation technique d'artefacts esthétiquement fonctionnels remettant en cause les normes culturelles régissant l'informatique et la musique<br>Music is a cultural and creative practice that enables humans to express a variety of feelings and intentions through sound. Machine learning opens many prospects for designing human expression in interactive music systems. Yet, as a Computer Science discipline, machine learning remains mostly studied from an engineering sciences perspective, which often exclude humans and musical interaction from the loop of the created systems. In this dissertation, I argue in favour of designing with machine learning for interactive music systems. I claim that machine learning must be first and foremost situated in human contexts to be researched and applied to the design of interactive music systems. I present four interdisciplinary studies that support this claim, using human-centred methods and model prototypes to design and apply machine learning to four situated musical tasks: motion-sound mapping, sonic exploration, synthesis exploration, and collective musical interaction. Through these studies, I show that model prototyping helps envision designs of machine learning with human users before engaging in model engineering. I also show that the final human-centred machine learning systems not only helps humans create static musical artifacts, but supports dynamic processes of expression between humans and machines. I call co-expression these processes of musical interaction between humans - who may have an expressive and creative impetus regardless of their expertise - and machines - whose learning abilities may be perceived as expressive by humans. In addition to these studies, I present five applications of the created model prototypes to the design of interactive music systems, which I publicly demonstrated in workshops, exhibitions, installations, and performances. Using a reflexive approach, I argue that the musical contributions enabled by such design practice with machine learning may ultimately complement the scientific contributions of human-centred machine learning. I claim that music research can thus be led through dispositif design, that is, through the technical realization of aesthetically-functioning artifacts that challenge cultural norms on computer science and music
2
Gosselin, Philippe-Henri. "Apprentissage interactif pour la recherche par le contenu dans les bases multimédias." Habilitation à diriger des recherches, Université de Cergy Pontoise, 2011. http://tel.archives-ouvertes.fr/tel-00660316.
Повний текст джерелаАнотація:
Les bases actuelles de données multimédia nécessitent des outils de plus en plus avancés pour pouvoir être parcourues avec efficacité. Dans ce contexte, la recherche en interaction avec un utilisateur est une approche qui permet de résoudre des requêtes à la sémantique complexe avec rapidité, sans pour autant nécessiter un haut niveau d'expertise utilisateur. Parmi les différents éléments intervenant dans la conception d'un système de recherche interactive, deux parties essentielles interviennent: l'indexation et la similarité entre les documents multimédia, et la gestion du processus interactif. Dans le contexte de la recherche multimédia par le contenu, on s'appuie sur des descriptions visuelles extraites automatiquement des documents. Suite à cette étape initiale, il est nécessaire de produire des structures de données, appelées index, ainsi qu'une métrique capable de comparer ces structures. Pour ce faire, nous proposons de représenter un document sous la forme d'un graphe, où chaque sommet du graphe représente une partie du document (région, point d'intérêt, ...) et chaque arête du graphe représente une relation entre deux parties du document. Puis, nous introduisons des métriques associées, sous la forme de fonctions noyaux sur graphes, qui permettent d'utiliser ces représentations complexes avec les méthodes d'apprentissages Hilbertiennes, telle que les SVMs. La gestion du processus interactif entre le système et un utilisateur a fait d'important progrès grâce à l'approche dite par apprentissage actif. Les premières approches proposent des critères pertinents pour la sélection de document à faire annoter par l'utilisateur, dans le but de trouver les documents recherchés au plus vite. Dans ce contexte, nous proposons d'aller plus loin en nous intéressant à la fabrication "en ligne" d'index et métriques associées en fonction de la nature de la recherche. En d'autres termes, nous proposons de remplacer le schéma traditionnel, où un unique index général est utilisé, par un schéma d'indexation active, où chaque utilisateur dispose d'un index dédié à sa requête.
3
Sungeelee, Vaynee. "Human-Machine Co-Learning : interactive curriculum generation for the acquisition of motor skills." Electronic Thesis or Diss., Sorbonne université, 2024. https://theses.hal.science/tel-04828514.
Повний текст джерелаАнотація:
L'acquisition de compétences motrices est le processus par lequel une personne est capable d'exécuter un mouvement avec plus de précision. Dans ce contexte, la pratique joue un rôle déterminant ; elle contribue à améliorer les performances de l'apprenant, mais n'y est pas toujours adapté. Un moyen de personnaliser l'apprentissage est de créer des séquences d'apprentissage qui conviennent à l'apprenant. Cependant, créer ces séquences manuellement demande du temps, ce qui rend cette démarche peu pratique. La génération automatique de séquences d'apprentissage peut remédier à ce problème. Les stratégies adaptatives qui structurent l'entraînement, c'est-à-dire la séquence de tâche à effectuer, en tenant compte de la difficulté de la tâche et du niveau de compétence de l'individu ont ainsi le potentiel d'améliorer l'apprentissage moteur. Ce double processus d'apprentissage par une machine de la séquence adaptée à un apprenant humain est ce que nous appelons le co-apprentissage. Dans cette thèse, nous abordons le co-apprentissage humain-machine dans le cas de l'apprentissage moteur, c'est-à-dire où les séquences d'apprentissage sont apprises en même temps que l'humain apprend à faire la tâche motrice.Les algorithmes de Machine Learning sont capables d'analyser les tendances d'apprentissage de chaque apprenant et ainsi adapter les consignes d'entraînement en fonction de ses besoins. Ils peuvent aussi être utilisés pour contrôler une interface homme-machine, dans un contexte où l'humain apprend à adapter ses mouvements (par exemple, pour apprendre à contrôler une prothèse). Dans cette thèse, le Machine Learning est utilisé afin de faciliter l'acquisition d'habiletés motrices. Cependant, une approche fondée sur le Machine Learning se heurte à de nombreuses exigences : (i) peu de données sont disponibles pour entraîner les algorithmes, (ii) la nature interactive du système implique un entraînement de courte durée (iii) l'efficacité des algorithmes dépend d'une évaluation précise du niveau de compétence de l'apprenant, qui est difficile à mesurer en pratique et (iv) le degré de contrôle que l'humain exerce sur l'entraînement du modèle de Machine Learning peut impacter comment l'humain apprend la tâche et construit un modèle mental lui permettant de prévoir le comportement du système.Les objectifs de cette thèse s'articulent autour de deux axes principaux:(i) élaborer une stratégie pour structurer l'apprentissage de tâches motrices (ii)) étudier les systèmes interactifs capables de s'adapter à, mais aussi d'être instruits par l'apprenant, pour optimiser les performances motrices. À travers deux études, nous explorons différentes stratégies d'élaboration de programmes d'apprentissage pour effectuer des tâches motrices. Dans la première étude, nous évaluons la précision et la fluidité du mouvement lors de l'exécution d'une tâche visuo-motrice. La seconde étude explore le contrôle de prothèse fondé sur la technique de reconnaissance de formes. Elle évalue à la fois la précision de reconnaissance des gestes par un algorithme de Machine Learning et la compréhension du système par les participants.Nos résultats contribuent aux travaux dans le domaine de l'apprentissage adaptatif des habiletés motrices ainsi qu'à l'interaction humain-machine. Ils démontrent qu'adapter les tâches motrices à l'apprenant présente des avantages en termes de performance des participants et de leur compréhension du système. Ces résultats offrent des perspectives pour créer des protocoles d'entraînement et faciliter leur implémentation dans un contexte appliqué<br>Motor skill acquisition is the process by which someone is able to perform a movement more accurately. In this context, practice plays a crucial role. However, practice is not always adapted to each learner's needs and learning journey. Generating personalised instructions manually is time-consuming and therefore impractical. Creating personalized practice sessions automatically is one way to alleviate this problem. Adaptive strategies that structure training, i.e. , the sequence of tasks executed, according to task difficulty and skill level have the potential to improve motor learning for the individual. This dual process of a machine learning the sequence adapted to a human learner and the human learning from the machine, is what we call co-learning. In this thesis, we study human-machine co-learning in the context of motor learning, i.e., learning sequences are generated at the same time as the human learns to perform the motor task.Machine learning algorithms can analyze the learning tendencies of individual learners and adapt training instructions accordingly. They can also be used to control a human-machine interface, during which humans learn to adapt their movements (e.g. prosthesis control). In this thesis, we leverage Machine Learning to facilitate the acquisition of motor skills. However, the use of Machine Learning to achieve this goal involves challenges : (i) few data is available to train the algorithms, (ii) the interactive nature of the system requires rapid training of machine learning algorithms. (iii) the effectiveness of the algorithms depends on a precise assessment of the learner's skill level, which is difficult to measure in practice and (iv) the degree of control provided to humans when training the machine learning model can impact their learning and the way they build a mental model to predict the system's behavior.The aims of this thesis are twofold: (i) to develop a strategy for structuring the learning of motor tasks (ii) to study interactive systems that can adapt to and be adapted by the learner to provide guidance during practice. Through two studies, we explore different strategies to sequence motor learning tasks. In the first study, we evaluate the accuracy and smoothness of movement execution during the performance of a visuo-motor task. The second study explores how to train a machine learning algorithm in a prosthesis control task. We evaluate both the recognition accuracy of gestures provided by participants as well as participants' understanding of the system.Our results contribute to the field of adaptive learning of motor skills and Human-Computer Interaction. They demonstrate that adapting motor tasks to the learner has advantages in terms of participants' performance and understanding of the task. These results provide insights for creating training protocols and facilitating their transition to applied contexts
4
Crochepierre, Laure. "Apprentissage automatique interactif pour les opérateurs du réseau électrique." Electronic Thesis or Diss., Université de Lorraine, 2022. http://www.theses.fr/2022LORR0112.
Повний текст джерелаАнотація:
Dans le contexte de la transition énergétique et de l'augmentation des interconnexions entre les réseaux de transport d'électricité en Europe, les opérateurs du réseau français doivent désormais faire face à davantage de fluctuations et des dynamiques nouvelles sur le réseau. Pour garantir la sûreté de ce réseau, les opérateurs s'appuient sur des logiciels informatiques permettant de réaliser des simulations, ou de suivre l'évolution d'indicateurs créés manuellement par des experts grâce à leur connaissance du fonctionnement du réseau. Le gestionnaire de réseau de transport d'électricité français RTE (Réseau de Transport d'Electricité) s'intéresse notamment aux développements d'outils permettant d'assister les opérateurs dans leur tâche de surveillance des transits sur les lignes électriques. Les transits sont en effet des grandeurs particulièrement importantes pour maintenir le réseau dans un état de sécurité, garantissant la sûreté du matériel et des personnes. Cependant, les indicateurs utilisés ne sont pas faciles à mettre à jour du fait de l'expertise nécessaire pour les construire et les analyser. Pour répondre à la problématique énoncée, cette thèse a pour objet la construction d'indicateurs, sous la forme d'expressions symboliques, permettant d'estimer les transits sur les lignes électriques. Le problème est étudié sous l'angle de la Régression Symbolique et investigué à la fois par des approches génétiques d'Evolution Grammaticale et d'Apprentissage par Renforcement dans lesquelles la connaissance experte, explicite et implicite, est prise en compte. Les connaissances explicites sur la physique et l'expertise du domaine électrique sont représentées sous la forme d'une grammaire non-contextuelle délimitant l'espace fonctionnel à partir duquel l'expression est créée. Une première approche d'Evolution Grammaticale Interactive propose d’améliorer incrémentalement les expressions trouvées par la mise à jour d'une grammaire entre les apprentissages évolutionnaires. Les expressions obtenues sur des données réelles issues de l'historique du réseau sont validées par une évaluation de métriques d'apprentissages, complétée par une évaluation de leur interprétabilité. Dans un second temps, nous proposons une approche par renforcement pour chercher dans un espace délimité par une grammaire non-contextuelle afin de construire une expression symbolique pertinente pour des applications comportant des contraintes physiques. Cette méthode est validée sur des données de l'état de l'art de la régression symbolique, ainsi qu’un jeu de données comportant des contraintes physiques pour en évaluer l'interprétabilité. De plus, afin de tirer parti des complémentarités entre les capacités des algorithmes d'apprentissage automatique et de l'expertise des opérateurs du réseau, des algorithmes interactifs de Régression Symbolique sont proposés et intégrés dans des plateformes interactives. L'interactivité est employée à la fois pour mettre à jour la connaissance représentée sous forme grammaticale, analyser, interagir avec et commenter les solutions proposées par les différentes approches. Ces algorithmes et interfaces interactifs ont également pour but de prendre en compte de la connaissance implicite, plus difficile à formaliser, grâce à l'utilisation de mécanismes d'interactions basés sur des suggestions et des préférences de l’utilisateur<br>In the energy transition context and the increase in interconnections between the electricity transmission networks in Europe, the French network operators must now deal with more fluctuations and new network dynamics. To guarantee the safety of the network, operators rely on computer software that allows them to carry out simulations or to monitor the evolution of indicators created manually by experts, thanks to their knowledge of the operation of the network. The French electricity transmission network operator RTE (Réseau de Transport d'Electricité) is particularly interested in developing tools to assist operators in monitoring flows on power lines. Flows are notably important to maintain the network in a safe state, guaranteeing the safety of equipment and people. However, the indicators used are not easy to update because of the expertise required to construct and analyze them.In order to address the stated problem, this thesis aims at constructing indicators, in the form of symbolic expressions, to estimate flows on power lines. The problem is studied from the Symbolic Regression perspective and investigated using both Grammatical Evolution and Reinforcement Learning approaches in which explicit and implicit expert knowledge is taken into account. Explicit knowledge about the physics and expertise of the electrical domain is represented in the form of a Context-Free Grammar to limit the functional space from which an expression is created. A first approach of Interactive Grammatical Evolution proposes to incrementally improve found expressions by updating a grammar between evolutionary learnings. Expressions are obtained on real-world data from the network history, validated by an analysis of learning metrics and an interpretability evaluation. Secondly, we propose a reinforcement approach to search in a space delimited by a Context-Free Grammar in order to build a relevant symbolic expression to applications involving physical constraints. This method is validated on state-of-the-art Symbolic Regression benchmarks and also on a dataset with physical constraints to assess its interpretability.Furthermore, in order to take advantage of the complementarities between the capacities of machine learning algorithms and the expertise of network operators, interactive Symbolic Regression algorithms are proposed and integrated into interactive platforms. Interactivity allows updating the knowledge represented in grammatical form and analyzing, interacting with, and commenting on the solutions found by the different approaches. These algorithms and interactive interfaces also aim to take into account implicit knowledge, which is more difficult to formalize, through interaction mechanisms based on suggestions and user preferences
5
Lai, Hien Phuong. "Vers un système interactif de structuration des index pour une recherche par le contenu dans des grandes bases d'images." Phd thesis, Université de La Rochelle, 2013. http://tel.archives-ouvertes.fr/tel-00934842.
Повний текст джерелаАнотація:
Cette thèse s'inscrit dans la problématique de l'indexation et la recherche d'images par le contenu dans des bases d'images volumineuses. Les systèmes traditionnels de recherche d'images par le contenu se composent généralement de trois étapes: l'indexation, la structuration et la recherche. Dans le cadre de cette thèse, nous nous intéressons plus particulièrement à l'étape de structuration qui vise à organiser, dans une structure de données, les signatures visuelles des images extraites dans la phase d'indexation afin de faciliter, d'accélérer et d'améliorer les résultats de la recherche ultérieure. A la place des méthodes traditionnelles de structuration, nous étudions les méthodes de regroupement des données (clustering) qui ont pour but d'organiser les signatures en groupes d'objets homogènes (clusters), sans aucune contrainte sur la taille des clusters, en se basant sur la similarité entre eux. Afin de combler le fossé sémantique entre les concepts de haut niveau sémantique exprimés par l'utilisateur et les signatures de bas niveau sémantique extraites automatiquement dans la phase d'indexation, nous proposons d'impliquer l'utilisateur dans la phase de clustering pour qu'il puisse interagir avec le système afin d'améliorer les résultats du clustering, et donc améliorer les résultats de la recherche ultérieure. En vue d'impliquer l'utilisateur dans la phase de clustering, nous proposons un nouveau modèle de clustering semi-supervisé interactif en utilisant les contraintes par paires (must-link et cannot-link) entre les groupes d'images. Tout d'abord, les images sont regroupées par le clustering non supervisé BIRCH (Zhang et al., 1996). Ensuite, l'utilisateur est impliqué dans la boucle d'interaction afin d'aider le clustering. Pour chaque itération interactive, l'utilisateur visualise les résultats de clustering et fournit des retours au système via notre interface interactive. Par des simples cliques, l'utilisateur peut spécifier les images positives ainsi que les images négatives pour chaque cluster. Il peut aussi glisser les images entre les clusters pour demander de changer l'affectation aux clusters des images. Les contraintes par paires sont ensuite déduites en se basant sur les retours de l'utilisateur ainsi que les informations de voisinage. En tenant compte de ces contraintes, le système réorganise les clusters en utilisant la méthode de clustering semi-supervisé proposée dans cette thèse. La boucle d'interaction peut être répétée jusqu'à ce que le résultat du clustering satisfasse l'utilisateur. Différentes stratégies pour déduire les contraintes par paires entre les images sont proposées. Ces stratégies sont analysées théoriquement et expérimentalement. Afin d'éviter que les résultats expérimentaux dépendent subjectivement de l'utilisateur humain, un agent logiciel simulant le comportement de l'utilisateur humain pour donner des retours est utilisé pour nos expérimentations. En comparant notre méthode avec la méthode de clustering semi-supervisé la plus populaire HMRF-kmeans (Basu et al., 2004), notre méthode donne de meilleurs résultats.
6
Pace, Aaron J. "Guided Interactive Machine Learning." Diss., CLICK HERE for online access, 2006. http://contentdm.lib.byu.edu/ETD/image/etd1355.pdf.
Повний текст джерела
7
Krishna, Sooraj. "Modelling communicative behaviours for different roles of pedagogical agents." Electronic Thesis or Diss., Sorbonne université, 2021. http://www.theses.fr/2021SORUS286.
Повний текст джерелаАнотація:
Les agents dans un environnement d'apprentissage peuvent avoir divers rôles et comportements sociaux qui peuvent influencer les objectifs et la motivation des apprenants de différentes manières. L'apprentissage autorégulé (SRL) est un cadre conceptuel complet qui englobe les aspects cognitifs, métacognitifs, comportementaux, motivationnels et affectifs de l'apprentissage et implique les processus de définition d'objectifs, de suivi des progrès, d'analyse des commentaires, d'ajustement des objectifs et des actions de l'apprenant. Dans cette thèse, nous présentons une interaction d'apprentissage multi-agent impliquant divers rôles d'agent pédagogique visant à améliorer l'autorégulation de l'apprenant tout en s'engageant dans une activité d'apprentissage socialement partagée. Nous avons utilisé des rôles distincts d'agents, définis par leurs attitudes sociales et leurs compétences, pour proposer des stratégies d'échafaudage de régulation spécifiques à l'apprenant. La méthodologie suivie dans cette thèse a commencé par la définition de rôles d'agent pédagogique dans un contexte de régulation socialement partagé et le développement d'une tâche d'apprentissage collaboratif pour faciliter l'autorégulation. Une série d'études d'utilisateurs a été menée pour comprendre les perceptions des apprenants sur les rôles des agents, les comportements associés et la tâche d'apprentissage. Dans l'ensemble, les travaux présentés dans cette thèse explorent comment divers rôles d'agents peuvent être utilisés pour fournir un échafaudage de régulation aux apprenants dans un contexte d'apprentissage socialement partagé<br>Agents in a learning environment can have various roles and social behaviours that can influence the goals and motivation of the learners in distinct ways. Self-regulated learning (SRL) is a comprehensive conceptual framework that encapsulates the cognitive, metacognitive, behavioural, motivational and affective aspects of learning and entails the processes of goal setting, monitoring progress, analyzing feedback, adjustment of goals and actions by the learner. In this thesis, we present a multi-agent learning interaction involving various pedagogical agent roles aiming to improve the self-regulation of the learner while engaging in a socially shared learning activity. We used distinct roles of agents, defined by their social attitudes and competence characteristics, to deliver specific regulation scaffolding strategies for the learner. The methodology followed in this Thesis started with the definition of pedagogical agent roles in a socially shared regulation context and the development of a collaborative learning task to facilitate self-regulation. Based on the learning task framework, we proposed a shared learning interaction consisting of a tutor agent providing external regulation support focusing on the performance of the learner and a peer agent demonstrating co-regulation strategies to promote self-regulation in the learner. A series of user studies have been conducted to understand the learner perceptions about the agent roles, related behaviours and the learning task. Altogether, the work presented in this thesis explores how various roles of agents can be utilised in providing regulation scaffolding to the learners in a socially shared learning context
8
Schild, Erwan. "De l’importance de valoriser l’expertise humaine dans l’annotation : application à la modélisation de textes en intentions à l’aide d’un clustering interactif." Electronic Thesis or Diss., Université de Lorraine, 2024. http://www.theses.fr/2024LORR0024.
Повний текст джерелаАнотація:
La tâche d'annotation, nécessaire à l'entraînement d'assistants conversationnels, fait habituellement appel à des experts du domaine à modéliser. Toutefois, l'annotation de données est connue pour être une tâche difficile en raison de sa complexité et sa subjectivité : elle nécessite par conséquent de solides compétences analytiques dans le but de modéliser les textes en intention de dialogue. De ce fait, la plupart des projets d'annotation choisissent de former les experts aux tâches d'analyse pour en faire des "super-experts". Dans cette thèse, nous avons plutôt décidé mettre l'accent sur les connaissances réelles des experts en proposant une nouvelle méthode d'annotation basée sur un Clustering Interactif. Celle-ci se base sur une coopération Homme/Machine, où la machine réalise un clustering pour proposer une base initiale d'apprentissage, et où l'expert annote des contraintes MUST-LINK ou CANNOT-LINK entre les données pour affiner itérativement la base d'apprentissage proposée. Une telle annotation présente l'avantage d'être plus instinctive, car les experts peuvent associer ou différencier les données en fonction de la similarité de leur cas d'usage, permettant ainsi de traiter les données comme ils le feraient professionnellement au quotidien. Au cours de nos études, nous avons pu montrer que cette méthode diminuait sensiblement la complexité de conception d'une base d'apprentissage, réduisant notamment la nécessité de formation des experts intervenant dans un projet d'annotation. Nous proposons une implémentation technique de cette méthode (algorithmes et interface graphique associée), ainsi qu'une étude des paramètres optimaux pour obtenir une base d'apprentissage cohérente en un minimum d'annotation. Nous réalisons également une étude de coûts (techniques et humains) permettant de confirmer que l'utilisation d'une telle méthode est réaliste dans un cadre industriel. De plus, afin que la méthode atteigne son plein potentiel, nous fournissons un ensemble de conseils, notamment : (1) des recommandations visant à cadrer la stratégie d'annotation, (2) une aide à l'identification et à la résolution des divergences d'opinion entre annotateurs, (3) des indicateurs de rentabilité pour chaque intervention de l'expert, et (4) des méthodes d'analyse de la pertinence de la base d'apprentissage en cours de construction. En conclusion, cette thèse offre une approche innovante pour concevoir une base d'apprentissage d'un assistant conversationnel, permettant d'impliquer les experts du domaine métier pour leurs vraies connaissances, tout en leur demandant un minimum de compétences analytiques et techniques. Ces travaux ouvrent ainsi la voie à des méthodes plus accessibles pour la construction de ces assistants<br>Usually, the task of annotation, used to train conversational assistants, relies on domain experts who understand the subject matter to model. However, data annotation is known to be a challenging task due to its complexity and subjectivity. Therefore, it requires strong analytical skills to model the text in dialogue intention. As a result, most annotation projects choose to train experts in analytical tasks to turn them into "super-experts". In this thesis, we decided instead to focus on the real knowledge of experts by proposing a new annotation method based on Interactive Clustering. This method involves a Human-Machine cooperation, where the machine performs clustering to provide an initial learning base, and the expert annotates MUST-LINK or CANNOT-LINK constraints between the data to iteratively refine the proposed learning base. Such annotation has the advantage of being more instinctive, as experts can associate or differentiate data according to the similarity of their use cases, allowing them to handle the data as they would professionally do on a daily basis. During our studies, we have been able to show that this method significantly reduces the complexity of designing a learning base, notably by reducing the need for training the experts involved in an annotation project. We provide a technical implementation of this method (algorithms and associated graphical interface), as well as a study of optimal parameters to achieve a coherent learning base with minimal annotation. We have also conducted a cost study (both technical and human) to confirm that the use of such a method is realistic in an industrial context. Finally, we provide a set of recommendations to help this method reach its full potential, including: (1) advice aimed at framing the annotation strategy, (2) assistance in identifying and resolving differences of opinion between annotators, (3) rentability indicators for each expert intervention, and (4) methods for analyzing the relevance of the learning base under construction. In conclusion, this thesis provides an innovative approach to design a learning base for a conversational assistant, involving domain experts for their actual knowledge, while requiring a minimum of analytical and technical skills. This work opens the way for more accessible methods for building such assistants
9
Georgiev, Nikolay. "Assisting physiotherapists by designing a system utilising Interactive Machine Learning." Thesis, Uppsala universitet, Institutionen för informatik och media, 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-447489.
Повний текст джерелаАнотація:
Millions of people throughout the world suffer from physical injuries and impairments and require physiotherapy to successfully recover. There are numerous obstacles in the way of having access to the necessary care – high costs, shortage of medical personnel and the need to travel to the appropriate medical facilities, something even more challenging during the Covid-19 pandemic. One approach to addressing this issue is to incorporate technology in the practice of physiotherapists, allowing them to help more patients. Using research through design, this thesis explores how interactive machine learning can be utilised in a system, designed for aiding physiotherapists. To this end, after a literature review, an informal case study was conducted. In order to explore what functionality the suggested system would need, an interface prototype was iteratively developed and subsequently evaluated through formative testing by three physiotherapists. All participants found value in the proposed system, and were interested in how such a system can be implemented and potentially used in practice. In particular the ability of the system to monitor the correct execution of the exercises by the patient, and the increased engagement during rehabilitative training brought by the sonification. Several suggestions for future developments in the topic are also presented at the end of this work.
10
Kim, Been. "Interactive and interpretable machine learning models for human machine collaboration." Thesis, Massachusetts Institute of Technology, 2015. http://hdl.handle.net/1721.1/98680.
Повний текст джерелаАнотація:
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Aeronautics and Astronautics, 2015.<br>Cataloged from PDF version of thesis.<br>Includes bibliographical references (pages 135-143).<br>I envision a system that enables successful collaborations between humans and machine learning models by harnessing the relative strength to accomplish what neither can do alone. Machine learning techniques and humans have skills that complement each other - machine learning techniques are good at computation on data at the lowest level of granularity, whereas people are better at abstracting knowledge from their experience, and transferring the knowledge across domains. The goal of this thesis is to develop a framework for human-in-the-loop machine learning that enables people to interact effectively with machine learning models to make better decisions, without requiring in-depth knowledge about machine learning techniques. Many of us interact with machine learning systems everyday. Systems that mine data for product recommendations, for example, are ubiquitous. However these systems compute their output without end-user involvement, and there are typically no life or death consequences in the case the machine learning result is not acceptable to the user. In contrast, domains where decisions can have serious consequences (e.g., emergency response panning, medical decision-making), require the incorporation of human experts' domain knowledge. These systems also must be transparent to earn experts' trust and be adopted in their workflow. The challenge addressed in this thesis is that traditional machine learning systems are not designed to extract domain experts' knowledge from natural workflow, or to provide pathways for the human domain expert to directly interact with the algorithm to interject their knowledge or to better understand the system output. For machine learning systems to make a real-world impact in these important domains, these systems must be able to communicate with highly skilled human experts to leverage their judgment and expertise, and share useful information or patterns from the data. In this thesis, I bridge this gap by building human-in-the-loop machine learning models and systems that compute and communicate machine learning results in ways that are compatible with the human decision-making process, and that can readily incorporate human experts' domain knowledge. I start by building a machine learning model that infers human teams' planning decisions from the structured form of natural language of team meetings. I show that the model can infer a human teams' final plan with 86% accuracy on average. I then design an interpretable machine learning model then "makes sense to humans" by exploring and communicating patterns and structure in data to support human decision-making. Through human subject experiments, I show that this interpretable machine learning model offers statistically significant quantitative improvements in interpretability while preserving clustering performance. Finally, I design a machine learning model that supports transparent interaction with humans without requiring that a user has expert knowledge of machine learning technique. I build a human-in-the-loop machine learning system that incorporates human feedback and communicates its internal states to humans, using an intuitive medium for interaction with the machine learning model. I demonstrate the application of this model for an educational domain in which teachers cluster programming assignments to streamline the grading process.<br>by Been Kim.<br>Ph. D.
11
Sahoo, Shibashankar. "Soft machine : A pattern language for interacting with machine learning algorithms." Thesis, Umeå universitet, Designhögskolan vid Umeå universitet, 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-182467.
Повний текст джерелаАнотація:
The computational nature of soft computing e.g. machine learning and AI systems have been hidden by seamless interfaces for almost two decades now. It has led to the loss of control, inability to explore, and adapt to needs and privacy at an individual level to social-technical problems on a global scale. I propose a soft machine - a set of cohesive design patterns or ‘seams’ to interact with everyday ‘black-box’ algorithms. Through participatory design and tangible sketching, I illustrate several interaction techniques to show how people can naturally control, explore, and adapt in-context algorithmic systems. Unlike existing design approaches, I treat machine learning as playful ‘design material’ finding moments of interplay between human common sense and statical intelligence. Further, I conceive machine learning not as a ‘technology’ but rather as an iterative training ‘process’, which eventually changes the role of user from a passive consumer of technology to an active trainer of algorithms.
12
Lagerkvist, Love. "Neural Novelty — How Machine Learning Does Interactive Generative Literature." Thesis, Malmö universitet, Fakulteten för kultur och samhälle (KS), 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:mau:diva-21222.
Повний текст джерелаАнотація:
Every day, machine learning (ML) and artificial intelligence (AI) embeds itself further into domestic and industrial technologies. Interaction de- signers have historically struggled to engage directly with the subject, facing a shortage of appropriate methods and abstractions. There is a need to find ways though which interaction design practitioners might integrate ML into their work, in order to democratize and diversify the field. This thesis proposes a mode of inquiry that considers the inter- active qualities of what machine learning does, as opposed the tech- nical specifications of what machine learning is. A shift in focus from the technicality of ML to the artifacts it creates allows the interaction designer to situate its existing skill set, affording it to engage with ma- chine learning as a design material. A Research-through-Design pro- cess explores different methodological adaptions, evaluated through user feedback and an-in depth case analysis. An elaborated design experiment, Multiverse, examines the novel, non-anthropomorphic aesthetic qualities of generative literature. It prototypes interactions with bidirectional literature and studies how these transform the reader into a cybertextual “user-reader”. The thesis ends with a discussion on the implications of machine written literature and proposes a number of future investigations into the research space unfolded through the prototype.
13
Gogia, Sumit. "Insight : interactive machine learning for complex graphics selection." Thesis, Massachusetts Institute of Technology, 2016. http://hdl.handle.net/1721.1/106021.
Повний текст джерелаАнотація:
Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2016.<br>This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.<br>Cataloged from student-submitted PDF version of thesis.<br>Includes bibliographical references (pages 89-91).<br>Modern vector graphics editors support the creation of a wonderful variety of complex designs and artwork. Users produce highly realistic illustrations, stylized representational art, even nuanced data visualizations. In light of these complex graphics, selections, representations of sets of objects that users want to manipulate, become more complex as well. Direct manipulation tools that artists and designers find accessible and useful for editing graphics such as logos and icons do not have the same applicability in these more complex cases. Given that selection is the first step for nearly all editing in graphics, it is important to enable artists and designers to express these complex selections. This thesis explores the use of interactive machine learning techniques to improve direct selection interfaces. To investigate this approach, I created Insight, an interactive machine learning selection tool for making a relevant class of complex selections: visually similar objects. To make a selection, users iteratively provide examples of selection objects by clicking on them in the graphic. Insight infers a selection from the examples at each step, allowing users to quickly understand results of actions and reactively shape the complex selection. The interaction resembles the direct manipulation interactions artists and designers have found accessible, while helping express complex selections by inferring many parameter changes from simple actions. I evaluated Insight in a user study of digital designers and artists, finding that Insight enabled users to effectively and easily make complex selections not supported by state-of-the-art vector graphics editors. My results contribute to existing work by both indicating a useful approach for providing complex representation access to artists and designers, and showing a new application for interactive machine learning.<br>by Sumit Gogia.<br>M. Eng.
14
Bouillon, Manuel. "Apprentissage actif en-ligne d'un classifieur évolutif, application à la reconnaissance de commandes gestuelles." Thesis, Rennes, INSA, 2016. http://www.theses.fr/2016ISAR0019/document.
Повний текст джерелаАнотація:
L'utilisation de commandes gestuelles est une nouvelle méthode d'interaction sur interface tactile. Une bonne méthode pour faciliter la mémorisation de ces commandes gestuelles est de laisser l'utilisateur les personnaliser. Ce contexte applicatif induit une situation d'apprentissage croisé, où l'utilisateur doit mémoriser le jeu de symboles elle système doit apprendre à reconnaître les différents symboles. Cela implique un certain nombre de contraintes, à la fois sur le système de reconnaissance de symboles ct sur le système de supervision de son apprentissage. Il faut par exemple que le classifieur puisse apprendre à partir de peu de données, continuer à apprendre pendant son utilisation et suivre toute évolution des données indéfiniment. Le superviseur doit quant à lui optimiser la coopération entre l'utilisateur et le système de reconnaissance pour minimiser les interactions tout en maximisant l'apprentissage. Cette thèse présente d'une part, le système d'apprentissage évolutif Evolve oo, capable d'apprendre rapidement il partir de peu de données et de suivre les changements de concepts. D'autre part, elle introduit le superviseur actif en-ligne lntuiSup qui permet d'optimiser la coopération entre le système et l'utilisateur, lors de l'utilisation de commandes gestuelles personnalisées notamment Evolve oo est un système d'inférence floue, capable d'apprendre rapidement grâce aux capacités génératrices des prémisses des règles, tout en permettant d'obtenir une précision élevée grâce aux capacités discriminantes des conclusions d'ordre un. L'intégration d'oubli dans le processus d'apprentissage permet de maintenir le gain de l'apprentissage indéfiniment, permettant ainsi l'ajout de classes à n'importe quel moment de l'utilisation du système ct garantissant son évolutivité « à vie». Le superviseur actif en-ligne lntuiSup permet d'optimiser les interactions avec l'utilisateur pour entraîner un système d'apprentissage lorsque l'utilisateur est dans la boucle. Il permet de faire évoluer la proportion de données que l'utilisateur doit étiqueter en fonction de la difficulté du problème et de l'évolution de l'environnement (changements de concepts). L'utilisation d'une méthode de« dopage» de l'apprentissage permet d'optimiser la répartition de ces interactions avec l'utilisateur pour maximiser leur impact sur l'apprentissage<br>Using gesture commands is a new way of interacting with touch sensitive interfaces. In order to facilitate user memorization of several commands, it is essential to let the user customize the gestures. This applicative context gives rise to a crosslearning situation, where the user has to memorize the set of commands and the system has to learn and recognize the different gestures. This situation implies several requirements, from the recognizer and from the system that supervizes its learning process. For instance, the recognizer has to be able to learn from few data samples, to keep learning during its use and to follow indefinitely any change of the data now. The supervisor has to optimize the cooperation between the recognizer and the system to minimize user interactions while maximizing recognizer learning. This thesis presents on the one hand the evolving recognition system Evolve oo, that is capable of fast teaming from few data samples, and that follows concept drifts. On the other hand, this thesis also presents the on line active supervisor lntuiSup, that optimizes user-system cooperation when the user is in the training loop, as during customized gesture command use for instance. The evolving classifier Evolve oo is a fuzzy inference system that is fast learning thanks to the generative capacity of rule premises, and at the same time giving high precision thanks to the discriminative capacity of first order rule conclusion. The use of forgetting in the learning process allows to maintain the learning gain indefinitely, enabling class adding at any stage of system learning, and guaranteeing lifelong evolving capacity. The on line active supervisor IntuiSup optimizes user interactions to train a classifier when the user is in the training loop. The proportion of data that is labeled by the user evolves to adapt to problem difficulty and to follow environment evolution (concept drift s). The use of a boosting method optimizes the timing of user interactions to maximize their impact on classifier learning process
15
Md, Noor Mohammad Faizuddin. "Machine learning techniques for implicit interaction using mobile sensors." Thesis, University of Glasgow, 2016. http://theses.gla.ac.uk/7723/.
Повний текст джерелаАнотація:
Interactions in mobile devices normally happen in an explicit manner, which means that they are initiated by the users. Yet, users are typically unaware that they also interact implicitly with their devices. For instance, our hand pose changes naturally when we type text messages. Whilst the touchscreen captures finger touches, hand movements during this interaction however are unused. If this implicit hand movement is observed, it can be used as additional information to support or to enhance the users’ text entry experience. This thesis investigates how implicit sensing can be used to improve existing, standard interaction technique qualities. In particular, this thesis looks into enhancing front-of-device interaction through back-of-device and hand movement implicit sensing. We propose the investigation through machine learning techniques. We look into problems on how sensor data via implicit sensing can be used to predict a certain aspect of an interaction. For instance, one of the questions that this thesis attempts to answer is whether hand movement during a touch targeting task correlates with the touch position. This is a complex relationship to understand but can be best explained through machine learning. Using machine learning as a tool, such correlation can be measured, quantified, understood and used to make predictions on future touch position. Furthermore, this thesis also evaluates the predictive power of the sensor data. We show this through a number of studies. In Chapter 5 we show that probabilistic modelling of sensor inputs and recorded touch locations can be used to predict the general area of future touches on touchscreen. In Chapter 7, using SVM classifiers, we show that data from implicit sensing from general mobile interactions is user-specific. This can be used to identify users implicitly. In Chapter 6, we also show that touch interaction errors can be detected from sensor data. In our experiment, we show that there are sufficient distinguishable patterns between normal interaction signals and signals that are strongly correlated with interaction error. In all studies, we show that performance gain can be achieved by combining sensor inputs.
16
Turner, Jonathan Milton. "Obstacle avoidance and path traversal using interactive machine learning /." Diss., CLICK HERE for online access, 2007. http://contentdm.lib.byu.edu/ETD/image/etd1905.pdf.
Повний текст джерела
17
Wood, David K. "Learning from Gross Motion Observations of Human-Machine Interaction." Thesis, The University of Sydney, 2011. https://hdl.handle.net/2123/29223.
Повний текст джерелаАнотація:
This thesis discusses the problems inherent in the modelling and classification of
human interactions with robots using gross motions observations. Contributions to
this field are one approach by which robots can be made socially aware, at a low
enough cost for the commercialisation of such systems to be viable. In general, it
cheaper and simpler both in terms of sensing requirements and computational power
to determine the position of a person participating in an interaction than to attempt
to perform more advanced operations such as face detection and recognition, gaze
tracking or gesture recognition. Being able to perform classification and modelling of
human behaviour from gross motion observations is a useful ability for the designers
of such HRI systems to have at their disposal.
Two contributions are made to the problem of gross motion modelling and classifica—
tion. The first is an approach to measuring error levels implicit to the models learned
in a generative classification scenario. By comparing the results from these model—
based error measures to the results obtained from more traditional data-based error
measures an assessment can be made about how well the internal models within the
classifier represent the true state of the world. A method is also presented to sum—
marise these comparisons using the symmetric Kullback—Leibler divergence, enabling
the rapid analysis of the large numbers of classifiers produced with the application of
cross—validation techniques.
The second contribution is a taxonomy of feature representations and a set of design
rules derived from this taxonomy for the representation of human—robot interaction
modelling features. These rules are focussed on gross motion features, but can be
extended to cover almost any human-robot interaction modelling or classification
task.
These two contributions are then demonstrated on interaction data gathered from the
Fish—Bird new media artwork. This is a challenging problem due to the interaction
parameters being modelled, however the use of a rigorous design approach and the
application of the divergence measures derived earlier in the thesis enable targeted
analysis and useful conclusions to be drawn. Results are shown to demonstrate these
applications.
18
Turner, Jonathan M. "Obstacle Avoidance and Path Traversal Using Interactive Machine Learning." BYU ScholarsArchive, 2007. https://scholarsarchive.byu.edu/etd/1006.
Повний текст джерелаАнотація:
Recently there has been a growing interest in using robots in activities that are dangerous or cost prohibitive for humans to do. Such activities include military uses and space exploration. While robotic hardware is often capable of being used in these types of situations, the ability of human operators to control robots in an effective manner is often limited. This deficiency is often related to the control interface of the robot and the level of autonomy that control system affords the human operator. This thesis describes a robot control system, called the safe/unsafe system, which gives a human operator the ability to quickly define how the system can cause the robot to automatically perform obstacle avoidance. This definition system uses interactive machine learning to ensure that the obstacle avoidance is both easy for a human operator to use and can perform well in different environments. Initial, real world tests show that system is effective at automatic obstacle avoidance.
19
Gustafson, Jonas. "Using Machine Learning to Identify Potential Problem Gamblers." Thesis, Umeå universitet, Institutionen för tillämpad fysik och elektronik, 2019. http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-163640.
Повний текст джерелаАнотація:
In modern casinos, personnel exist to advise, or in some cases, order individuals to stop gambling if they are found to be gambling in a destructive way, but what about online gamblers? This thesis evaluated the possibility of using machine learning as a supplement for personnel in real casinos when gambling online. This was done through supervised learning or more specifically, a decision tree algorithm called CART. Studies showed that the majority of problem gamblers would find it helpful to have their behavioral patterns collected to be able to identify their risk of becoming a problem gambler before their problem started. The collected behavioral features were time spent gambling, the rate of won and lost money and the number of deposits made, all these during a specific period of time. An API was implemented for casino platforms to connect to and give collected data about their users, and to receive responses to notify users about their situation. Unfortunately, there were no platforms available to test this on players gambling live. Therefore a web based survey was implemented to test if the API would work as expected. More studies could be conducted in this area, finding more features to convert for computers to understand and implement into the learning algorithm.
20
Chen, Si. "Active Learning Under Limited Interaction with Data Labeler." Thesis, Virginia Tech, 2021. http://hdl.handle.net/10919/104894.
Повний текст джерелаАнотація:
Active learning (AL) aims at reducing labeling effort by identifying the most valuable unlabeled data points from a large pool. Traditional AL frameworks have two limitations: First, they perform data selection in a multi-round manner, which is time-consuming and impractical. Second, they usually assume that there are a small amount of labeled data points available in the same domain as the data in the unlabeled pool.
In this thesis, we initiate the study of one-round active learning to solve the first issue. We propose DULO, a general framework for one-round setting based on the notion of data utility functions, which map a set of data points to some performance measure of the model trained on the set. We formulate the one-round active learning problem as data utility function maximization.
We then propose D²ULO on the basis of DULO as a solution that solves both issues. Specifically, D²ULO leverages the idea of domain adaptation (DA) to train a data utility model on source labeled data. The trained utility model can then be used to select high-utility data in the target domain and at the same time, provide an estimate for the utility of the selected data. Our experiments show that the proposed frameworks achieves better performance compared with state-of-the-art baselines in the same setting. Particularly, D²ULO is applicable to the scenario where the source and target labels have mismatches, which is not supported by the existing works.<br>M.S.<br>Machine Learning (ML) has achieved huge success in recent years. Machine Learning technologies such as recommendation system, speech recognition and image recognition play an important role on human daily life. This success mainly build upon the use of large amount of labeled data: Compared with traditional programming, a ML algorithm does not rely on explicit instructions from human; instead, it takes the data along with the label as input, and aims to learn a function that can correctly map data to the label space by itself. However, data labeling requires human effort and could be time-consuming and expensive especially for datasets that contain domain-specific knowledge (e.g., disease prediction etc.) Active Learning (AL) is one of the solution to reduce data labeling effort. Specifically, the learning algorithm actively selects data points that provide more information for the model, hence a better model can be achieved with less labeled data.
While traditional AL strategies do achieve good performance, it requires a small amount of labeled data as initialization and performs data selection in multi-round, which pose great challenge to its application, as there is no platform provide timely online interaction with data labeler and the interaction is often time inefficient. To deal with the limitations, we first propose DULO which a new setting of AL is studied: data selection is only allowed to be performed once. To further broaden the application of our method, we propose D²ULO which is built upon DULO and Domain Adaptation techniques to avoid the use of initial labeled data. Our experiments show that both of the proposed two frameworks achieve better performance compared with state-of-the-art baselines.
21
Holmberg, Lars. "Human In Command Machine Learning." Licentiate thesis, Malmö universitet, Malmö högskola, Institutionen för datavetenskap och medieteknik (DVMT), 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:mau:diva-42576.
Повний текст джерелаАнотація:
Machine Learning (ML) and Artificial Intelligence (AI) impact many aspects of human life, from recommending a significant other to assist the search for extraterrestrial life. The area develops rapidly and exiting unexplored design spaces are constantly laid bare. The focus in this work is one of these areas; ML systems where decisions concerning ML model training, usage and selection of target domain lay in the hands of domain experts. This work is then on ML systems that function as a tool that augments and/or enhance human capabilities. The approach presented is denoted Human In Command ML (HIC-ML) systems. To enquire into this research domain design experiments of varying fidelity were used. Two of these experiments focus on augmenting human capabilities and targets the domains commuting and sorting batteries. One experiment focuses on enhancing human capabilities by identifying similar hand-painted plates. The experiments are used as illustrative examples to explore settings where domain experts potentially can: independently train an ML model and in an iterative fashion, interact with it and interpret and understand its decisions. HIC-ML should be seen as a governance principle that focuses on adding value and meaning to users. In this work, concrete application areas are presented and discussed. To open up for designing ML-based products for the area an abstract model for HIC-ML is constructed and design guidelines are proposed. In addition, terminology and abstractions useful when designing for explicability are presented by imposing structure and rigidity derived from scientific explanations. Together, this opens up for a contextual shift in ML and makes new application areas probable, areas that naturally couples the usage of AI technology to human virtues and potentially, as a consequence, can result in a democratisation of the usage and knowledge concerning this powerful technology.
22
Alcoverro, Vidal Marcel. "Stochastic optimization and interactive machine learning for human motion analysis." Doctoral thesis, Universitat Politècnica de Catalunya, 2014. http://hdl.handle.net/10803/285337.
Повний текст джерелаАнотація:
The analysis of human motion from visual data is a central issue in the computer vision research community as it enables a wide range of applications and it still remains a challenging problem when dealing with unconstrained scenarios and general conditions. Human motion analysis is used in the entertainment industry for movies or videogame production, in medical applications for rehabilitation or biomechanical studies. It is also used for human computer interaction in any kind of environment, and moreover, it is used for big data analysis from social networks such as Youtube or Flickr, to mention some of its use cases.
In this thesis we have studied human motion analysis techniques with a focus on its application for smart room environments. That is, we have studied methods that will support the analysis of people behavior in the room, allowing interaction with computers in a natural manner and in general, methods that introduce computers in human activity environments to enable new kind of services but in an unobstrusive mode. The thesis is structured in two parts, where we study the problem of 3D pose estimation from multiple views and the recognition of gestures using range sensors.
First, we propose a generic framework for hierarchically layered particle filtering (HPF) specially suited for motion capture tasks. Human motion capture problem generally involve tracking or optimization of high-dimensional state vectors where also one have to deal with multi-modal pdfs. HPF allow to overcome the problem by means of multiple passes through substate space variables. Then, based on the HPF framework, we propose a method to estimate the anthropometry of the subject, which at the end allows to obtain a human body model adjusted to the subject. Moreover, we introduce a new weighting function strategy for approximate partitioning of observations and a method that employs body part detections to improve particle propagation and weight evaluation, both integrated within the HPF framework.
The second part of this thesis is centered in the detection of gestures, and we have focused the problem of reducing annotation and training efforts required to train a specific gesture. In order to reduce the efforts required to train a gesture detector, we propose a solution based on online random forests that allows training in real-time, while receiving new data in sequence. The main aspect that makes the solution effective is the method we propose to collect the hard negatives examples while training the forests. The method uses the detector trained up to the current frame to test on that frame, and then collects samples based on the response of the detector such that they will be more relevant for training. In this manner, training is more effective in terms of the number of annotated frames required.<br>L'anàlisi del moviment humà a partir de dades visuals és un tema central en la recerca en visió per computador, per una banda perquè habilita un ampli espectre d'aplicacions i per altra perquè encara és un problema no resolt quan és aplicat en escenaris no controlats. L'analisi del moviment humà s'utilitza a l'indústria de l'entreteniment per la producció de pel·lícules i videojocs, en aplicacions mèdiques per rehabilitació o per estudis bio-mecànics. També s'utilitza en el camp de la interacció amb computadors o també per l'analisi de grans volums de dades de xarxes socials com Youtube o Flickr, per mencionar alguns exemples. En aquesta tesi s'han estudiat tècniques per l'anàlisi de moviment humà enfocant la seva aplicació en entorns de sales intel·ligents. És a dir, s'ha enfocat a mètodes que puguin permetre l'anàlisi del comportament de les persones a la sala, que permetin la interacció amb els dispositius d'una manera natural i, en general, mètodes que incorporin les computadores en espais on hi ha activitat de persones, per habilitar nous serveis de manera que no interfereixin en la activitat. A la primera part, es proposa un marc genèric per l'ús de filtres de partícules jeràrquics (HPF) especialment adequat per tasques de captura de moviment humà. La captura de moviment humà generalment implica seguiment i optimització de vectors d'estat de molt alta dimensió on a la vegada també s'han de tractar pdf's multi-modals. Els HPF permeten tractar aquest problema mitjançant multiples passades en subdivisions del vector d'estat. Basant-nos en el marc dels HPF, es proposa un mètode per estimar l'antropometria del subjecte, que a la vegada permet obtenir un model acurat del subjecte. També proposem dos nous mètodes per la captura de moviment humà. Per una banda, el APO es basa en una nova estratègia per les funcions de cost basada en la partició de les observacions. Per altra, el DD-HPF utilitza deteccions de parts del cos per millorar la propagació de partícules i l'avaluació de pesos. Ambdós mètodes són integrats dins el marc dels HPF. La segona part de la tesi es centra en la detecció de gestos, i s'ha enfocat en el problema de reduir els esforços d'anotació i entrenament requerits per entrenar un detector per un gest concret. Per tal de reduir els esforços requerits per entrenar un detector de gestos, proposem una solució basada en online random forests que permet l'entrenament en temps real, mentre es reben noves dades sequencialment. El principal aspecte que fa la solució efectiva és el mètode que proposem per obtenir mostres negatives rellevants, mentre s'entrenen els arbres de decisió. El mètode utilitza el detector entrenat fins al moment per recollir mostres basades en la resposta del detector, de manera que siguin més rellevants per l'entrenament. D'aquesta manera l'entrenament és més efectiu pel que fa al nombre de mostres anotades que es requereixen.
23
Patel, Vedang Vikrambhai. "Reduced Order Modeling For Fluid Structure Interaction Using Machine Learning." The Ohio State University, 2021. http://rave.ohiolink.edu/etdc/view?acc_num=osu1619276050317241.
Повний текст джерела
24
Xia, Guangyu. "Expressive Collaborative Music Performance via Machine Learning." Research Showcase @ CMU, 2016. http://repository.cmu.edu/dissertations/784.
Повний текст джерелаАнотація:
Techniques of Artificial Intelligence and Human-Computer Interaction have empowered computer music systems with the ability to perform with humans via a wide spectrum of applications. However, musical interaction between humans and machines is still far less musical than the interaction between humans since most systems lack any representation or capability of musical expression. This thesis contributes various techniques, especially machine-learning algorithms, to create artificial musicians that perform expressively and collaboratively with humans. The current system focuses on three aspects of expression in human-computer collaborative performance: 1) expressive timing and dynamics, 2) basic improvisation techniques, and 3) facial and body gestures. Timing and dynamics are the two most fundamental aspects of musical expression and also the main focus of this thesis. We model the expression of different musicians as co-evolving time series. Based on this representation, we develop a set of algorithms, including a sophisticated spectral learning method, to discover regularities of expressive musical interaction from rehearsals. Given a learned model, an artificial performer generates its own musical expression by interacting with a human performer given a predefined score. The results show that, with a small number of rehearsals, we can successfully apply machine learning to generate more expressive and human-like collaborative performance than the baseline automatic accompaniment algorithm. This is the first application of spectral learning in the field of music. Besides expressive timing and dynamics, we consider some basic improvisation techniques where musicians have the freedom to interpret pitches and rhythms. We developed a model that trains a different set of parameters for each individual measure and focus on the prediction of the number of chords and the number of notes per chord. Given the model prediction, an improvised score is decoded using nearest-neighbor search, which selects the training example whose parameters are closest to the estimation. Our result shows that our model generates more musical, interactive, and natural collaborative improvisation than a reasonable baseline based on mean estimation. Although not conventionally considered to be “music,” body and facial movements are also important aspects of musical expression. We study body and facial expressions using a humanoid saxophonist robot. We contribute the first algorithm to enable a robot to perform an accompaniment for a musician and react to human performance with gestural and facial expression. The current system uses rule-based performance-motion mapping and separates robot motions into three groups: finger motions, body movements, and eyebrow movements. We also conduct the first subjective evaluation of the joint effect of automatic accompaniment and robot expression. Our result shows robot embodiment and expression enable more musical, interactive, and engaging human-computer collaborative performance.
25
Willman, Martin. "Machine Learning to identify cheaters in online games." Thesis, Umeå universitet, Institutionen för tillämpad fysik och elektronik, 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-170973.
Повний текст джерелаАнотація:
Cheating in online games is a problem both on the esport stage and in the gaming community. When a player cheats, the competitors do not compete on the same terms anymore and this becomes a major problem when high price pools are involved in online games. In this master thesis, a machine learning approach will be developed and tested to try to identify cheaters in the first-person shooter game Counter-Strike : Global Offensive. The thesis will also go through how the game Counter-Strike : Global Offensive works, give examples of anti-cheat softwares that exists, analyse different cheats in the game, consider social aspects of cheating in online games, and give an introduction to machine learning. The machine learning approach was done by creating and evaluating a recurrent neural network with data from games played with the cheat aimbot and without the cheat aimbot. The recurrent neural network that was created in this master thesis should be considered as the first step towards creating a reliable anti-cheat machine learning algorithm. To possible increase the result of the recurrent neural network, more data and more data points from the game would be needed.
26
Chen, Li. "Searching for significant feature interaction from biological data." Diss., Online access via UMI:, 2007.
Знайти повний текст джерела
27
Charif, Omar. "Modelling and simulating individual's mobility : case study of Luxembourg and its greater region." Thesis, Compiègne, 2013. http://www.theses.fr/2013COMP2130.
Повний текст джерелаАнотація:
Dans le dernier siècle, le transport et en particulier l'utilisation des voitures privées a émergé comme une des sources principales d'émission de CO2 (deuxième derrière la production d'énergie). Plusieurs villes dans le monde ont mis en place des stratégies pour faire face à ce phénomène afin de limiter les impacts environnementaux néfastes. Certaines stratégies n'ont pas pu atteindre leur objectif, voire ils ont eu des réactions négatives auprès des individus. Le but de cette thèse de doctorat consiste à proposer une méthodologie et une plateforme pour la modélisation et la simulation du système de mobilité. Cette plateforme sera ensuite utilisée pour implémenter des scénarios d'occupation de sol et de transport dans un monde virtuel pour étudier leur impact sur le comportement humain en termes de mobilité. Pour la modélisation de la dynamique des déplacements évoqués par la mobilité locale, nous proposons une méthode hybride (automates cellulaires et systèmes multi-agents) permettant de traiter des données complexes tout en les intégrant au sein de différentes échelles spatio-temporelles<br>In the last century, transport and in particular the use of private cars has emerged as a major source of CO2 emissions (second behinf energy production). Several cities in the world have put in place strategies to deal with this problem and to reduce its adverse enviromental impacts. Some strategies could not achieve their objectives, and had negative reactions from individuals. The ail of this PhD thesis is to propose a methodology and a platform for modelling and simulating people mobility systems. The developed plat form is, then, used to implement land use and transportation scenarios and strategies in a virtual world to study their impact on human behavior in terms of mobility. To develop this platform, we propose a hybrid model, combining cellular automata and multi-agent systems, capable of handling the complexity of the mobility system able to present it at various spatial ans temporal scales
28
Stenström, Albin. "Clicking using the eyes, a machine learning approach." Thesis, Linköpings universitet, Interaktiva och kognitiva system, 2015. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-121834.
Повний текст джерелаАнотація:
This master thesis report describes the work of evaluating the approach of using an eye-tracker and machine learning to generate an interaction model for clicks. In the study, recordings were done from 10 participants using a quiz application, and machine learning was then applied. Models were created with varying quality from a machine learning view, although most models did not work well for interaction. One model was created that enable correct interaction 80\% of the time, although the specific circumstances for success were not identified. The conclusion of the thesis is that the approach works in some cases, but that more research needs to be done to evaluate general suitability, and approaches to make it work reliably.
29
Riviere, Jean-Philippe. "Capturing traces of the dance learning process." Electronic Thesis or Diss., université Paris-Saclay, 2020. http://www.theses.fr/2020UPASG054.
Повний текст джерелаАнотація:
Cette thèse porte sur la conception d’outils interactifs pour comprendre et faciliter l’apprentissage de la danse à partir de vidéos. Les processus d’apprentissage des danseurs représentent une source d’informations riches pour les chercheurs qui s’intéressent à la conception de systèmes soutenant l’apprentissage moteur. En effet, les danseurs experts réutilisent un large éventail de compétences qu’ils ont appris. Cependant, ces compétences sont en partie le résultat de connaissances implicites et incarnées,qui sont difficilement exprimables et verbalisables par un individu.Dans cette thèse, je soutiens que nous pouvons capturer et sauvegarder une trace des connaissances implicites des danseurs et les utiliser pour concevoir des outils interactifs qui soutiennent l’apprentissage de la danse. Mon approche consiste à étudier différentes sessions d’apprentissage de danse dans des contextes réels, aussi bien individuels que collaboratifs.Sur la base des résultats apportés par ces études, je contribue à une meilleure compréhension des processus implicites qui sous-tendent l’apprentissage de la danse dans des contextes individuels et collectifs. Je présente plusieurs stratégies d’apprentissage utilisées par des danseurs et j’affirme que l’on peut documenter ces stratégies en sauvegardant une trace de l’apprentissage. Je discute de l’opportunité que représente la capture de ces connaissances incarnées et j’apporte de nouvelles perspectives pour la conception d’outils d’aide à l’apprentissage du mouvement par la vidéo<br>This thesis focuses on designing interactive tools to understand and support dance learning from videos. Dancers' learning practice represents a rich source of information for researchers interested in designing systems that support motor learning. Indeed, dancers embody a wide range of skills that they reuse during new dance sequences learning. However, these skills are in part the result of embodied implicit knowledge. In this thesis, I argue that we can capture and save traces of dancers' embodied knowledge and use them to design interactive tools that support dance learning. My approach is to study real-life dance learning tasks in individual and collaborative settings. Based on the findings from all the studies, I discuss the challenge of capturing embodied knowledge to support dancers’ learning practice. My thesis highlights that although dancers’ learning processes are diverse, similar strategies emerge to structure their learning process. Finally, I bring and discuss new perspectives to the design of movement-based learning tools
30
Anderson, Corin R. "A machine learning approach to Web personalization /." Thesis, Connect to this title online; UW restricted, 2002. http://hdl.handle.net/1773/6875.
Повний текст джерела
31
Dias, Pedro Ricardo Gomes. "Recommending media content based on machine learning methods." Master's thesis, Faculdade de Ciências e Tecnologia, 2011. http://hdl.handle.net/10362/6581.
Повний текст джерелаАнотація:
Dissertação para obtenção do Grau de Mestre em
Engenharia Informática<br>Information is nowadays made available and consumed faster than ever before. This information technology generation has access to a tremendous deal of data and is left with
the heavy burden of choosing what is relevant. With the increasing growth of media
sources, the amount of content made available to users has become overwhelming and in need to be managed. Recommender systems emerged with the purpose of providing
personalized and meaningful content recommendations based on users’ preferences and usage history. Due to their utility and commercial potential, recommender systems integrate many audiovisual content providers and represent one of their most important and
valuable services. The goal of this thesis is to develop a recommender system based on
matrix factorization methods, capable of providing meaningful and personalized product
recommendations to individual users and groups of users, by taking into account
users’ rating patterns and biased tendencies, as well as their fluctuations throughout time.
32
Harbert, Christopher W. Shang Yi. "An application of machine learning techniques to interactive, constraint-based search." Diss., Columbia, Mo. : University of Missouri-Columbia, 2005. http://hdl.handle.net/10355/4324.
Повний текст джерелаАнотація:
Thesis (M.S.)--University of Missouri-Columbia, 2005.<br>The entire dissertation/thesis text is included in the research.pdf file; the official abstract appears in the short.pdf file (which also appears in the research.pdf); a non-technical general description, or public abstract, appears in the public.pdf file. Title from title screen of research.pdf file viewed on (December 12, 2006) Includes bibliographical references.
33
Parikh, Neena (Neena S. ). "Interactive tools for fantasy football analytics and predictions using machine learning." Thesis, Massachusetts Institute of Technology, 2015. http://hdl.handle.net/1721.1/100687.
Повний текст джерелаАнотація:
Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2015.<br>Cataloged from PDF version of thesis.<br>Includes bibliographical references (pages 83-84).<br>The focus of this project is multifaceted: we aim to construct robust predictive models to project the performance of individual football players, and we plan to integrate these projections into a web-based application for in-depth fantasy football analytics. Most existing statistical tools for the NFL are limited to the use of macro-level data; this research looks to explore statistics at a finer granularity. We explore various machine learning techniques to develop predictive models for different player positions including quarterbacks, running backs, wide receivers, tight ends, and kickers. We also develop an interactive interface that will assist fantasy football participants in making informed decisions when managing their fantasy teams. We hope that this research will not only result in a well-received and widely used application, but also help pave the way for a transformation in the field of football analytics.<br>by Neena Parikh.<br>M. Eng.
34
Westphal, Florian. "Efficient Document Image Binarization using Heterogeneous Computing and Interactive Machine Learning." Licentiate thesis, Blekinge Tekniska Högskola, Institutionen för datalogi och datorsystemteknik, 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:bth-16797.
Повний текст джерелаАнотація:
Large collections of historical document images have been collected by companies and government institutions for decades. More recently, these collections have been made available to a larger public via the Internet. However, to make accessing them truly useful, the contained images need to be made readable and searchable. One step in that direction is document image binarization, the separation of text foreground from page background. This separation makes the text shown in the document images easier to process by humans and other image processing algorithms alike. While reasonably well working binarization algorithms exist, it is not sufficient to just being able to perform the separation of foreground and background well. This separation also has to be achieved in an efficient manner, in terms of execution time, but also in terms of training data used by machine learning based methods. This is necessary to make binarization not only theoretically possible, but also practically viable. In this thesis, we explore different ways to achieve efficient binarization in terms of execution time by improving the implementation and the algorithm of a state-of-the-art binarization method. We find that parameter prediction, as well as mapping the algorithm onto the graphics processing unit (GPU) help to improve its execution performance. Furthermore, we propose a binarization algorithm based on recurrent neural networks and evaluate the choice of its design parameters with respect to their impact on execution time and binarization quality. Here, we identify a trade-off between binarization quality and execution performance based on the algorithm’s footprint size and show that dynamically weighted training loss tends to improve the binarization quality. Lastly, we address the problem of training data efficiency by evaluating the use of interactive machine learning for reducing the required amount of training data for our recurrent neural network based method. We show that user feedback can help to achieve better binarization quality with less training data and that visualized uncertainty helps to guide users to give more relevant feedback.<br>Scalable resource-efficient systems for big data analytics
35
Yu, Guoqiang. "Machine Learning to Interrogate High-throughput Genomic Data: Theory and Applications." Diss., Virginia Tech, 2011. http://hdl.handle.net/10919/28980.
Повний текст джерелаАнотація:
The missing heritability in genome-wide association studies (GWAS) is an intriguing open scientific problem which has attracted great recent interest. The interaction effects among risk factors, both genetic and environmental, are hypothesized to be one of the main missing heritability sources. Moreover, detection of multilocus interaction effect may also have great implications for revealing disease/biological mechanisms, for accurate risk prediction, personalized clinical management, and targeted drug design. However, current analysis of GWAS largely ignores interaction effects, partly due to the lack of tools that meet the statistical and computational challenges posed by taking into account interaction effects. Here, we propose a novel statistically-based framework (Significant Conditional Association) for systematically exploring, assessing significance, and detecting interaction effect. Further, our SCA work has also revealed new theoretical results and insights on interaction detection, as well as theoretical performance bounds. Using in silico data, we show that the new approach has detection power significantly better than that of peer methods, while controlling the running time within a permissible range. More importantly, we applied our methods on several real data sets, confirming well-validated interactions with more convincing evidence (generating smaller p-values and requiring fewer samples) than those obtained through conventional methods, eliminating inconsistent results in the original reports, and observing novel discoveries that are otherwise undetectable. The proposed methods provide a useful tool to mine new knowledge from existing GWAS and generate new hypotheses for further research.
Microarray gene expression studies provide new opportunities for the molecular characterization of heterogeneous diseases. Multiclass gene selection is an imperative task for identifying phenotype-associated mechanistic genes and achieving accurate diagnostic classification. Most existing multiclass gene selection methods heavily rely on the direct extension of two-class gene selection methods. However, simple extensions of binary discriminant analysis to multiclass gene selection are suboptimal and not well-matched to the unique characteristics of the multi-category classification problem. We report a simpler and yet more accurate strategy than previous works for multicategory classification of heterogeneous diseases. Our method selects the union of one-versus-everyone phenotypic up-regulated genes (OVEPUGs) and matches this gene selection with a one-versus-rest support vector machine. Our approach provides even-handed gene resources for discriminating both neighboring and well-separated classes, and intends to assure the statistical reproducibility and biological plausibility of the selected genes. We evaluated the fold changes of OVEPUGs and found that only a small number of high-ranked genes were required to achieve superior accuracy for multicategory classification. We tested the proposed OVEPUG method on six real microarray gene expression data sets (five public benchmarks and one in-house data set) and two simulation data sets, observing significantly improved performance with lower error rates, fewer marker genes, and higher performance sustainability, as compared to several widely-adopted gene selection and classification methods.<br>Ph. D.
36
Delaunay, Julien. "Explainability for machine learning models : from data adaptability to user perception." Electronic Thesis or Diss., Université de Rennes (2023-....), 2023. http://www.theses.fr/2023URENS076.
Повний текст джерелаАнотація:
Cette thèse se concentre sur la génération d'explications locales pour les modèles de machine learning déjà déployés, en recherchant les conditions optimales pour des explications pertinentes, prenant en compte à la fois les données et les besoins de l'utilisateur. L'objectif principal est de développer des méthodes produisant des explications pour n'importe quel modèle de prédiction, tout en veillant à ce que ces explications demeurent à la fois fidèles au modèle sous-jacent et compréhensibles par les utilisateurs qui les reçoivent. La thèse est divisée en deux parties. Dans la première, on améliore une méthode d'explication basée sur des règles. On introduit ensuite une approche pour évaluer l'adéquation des explications linéaires pour approximer un modèle à expliquer. Enfin, cette partie présente une expérimentation comparative entre deux familles de méthodes d'explication contrefactuelles, dans le but d'analyser les avantages de l'une par rapport à l'autre. La deuxième partie se concentre sur des expériences utilisateurs évaluant l'impact de trois méthodes d'explication et de deux représentations différentes. Ces expériences mesurent la perception en termes de compréhension et de confiance des utilisateurs en fonction des explications et de leurs représentations. L'ensemble de ces travaux contribue à une meilleure compréhension de la génération d'explications pour les modèles de machine learning, avec des implications potentielles pour l'amélioration de la transparence, de la confiance et de l'utilisabilité des systèmes d'IA déployés<br>This thesis explores the generation of local explanations for already deployed machine learning models, aiming to identify optimal conditions for producing meaningful explanations considering both data and user requirements. The primary goal is to develop methods for generating explanations for any model while ensuring that these explanations remain faithful to the underlying model and comprehensible to the users. The thesis is divided into two parts. The first enhances a widely used rule-based explanation method to improve the quality of explanations. It then introduces a novel approach for evaluating the suitability of linear explanations to approximate a model. Additionally, it conducts a comparative experiment between two families of counterfactual explanation methods to analyze the advantages of one over the other. The second part focuses on user experiments to assess the impact of three explanation methods and two distinct representations. These experiments measure how users perceive their interaction with the model in terms of understanding and trust, depending on the explanations and representations. This research contributes to a better explanation generation, with potential implications for enhancing the transparency, trustworthiness, and usability of deployed AI systems
37
Sullivan, Patrick Ryan. "ALJI: Active Listening Journal Interaction." Thesis, Virginia Tech, 2019. http://hdl.handle.net/10919/95207.
Повний текст джерелаАнотація:
Depression is a crippling burden on a great many people, and it is often well hidden. Mental health professionals are able to treat depression, but the general public is not well versed in recognizing depression symptoms or assessing their own mental health. Active Listening Journal Interaction (ALJI) is a computer program that seeks to identify and refer people suffering with depression to mental health support services. It does this through analyzing personal journal entries using machine learning, and then privately responding to the author with proper guidance. In this thesis, we focus on determining the feasibility and usefulness of the machine learning models that drive ALJI. With heavy data limitations, we cautiously report that with a single journal entry, our model detects when a person's symptoms warrant professional intervention with a 61% accuracy. A great amount of discussion on the proposed solution, methods, results, and future directions of ALJI is included.<br>Master of Science<br>An incredibly large number of people suffer from depression, and they can rightfully feel trapped or imprisoned by this illness. A very simple way to understand depression is to first imagine looking at the most beautiful sunset you've ever seen, and then imagine feeling absolutely nothing while looking that same sunset, and you can't explain why. When a person is depressed, they are likely to feel like a burden to those around them. This causes them to avoid social gathering and friends, making them isolated away from people that could support them. This worsens their depression and a terrible cycle begins. One of the best ways out of this cycle is to reveal the depression to a doctor or psychologist, and to ask them for guidance. However, many people don't see or realize this excellent option is open to them, and will continue to suffer with depression for far longer than needed.
This thesis describes an idea called the Active Listening Journal Interaction, or ALJI. ALJI acts just like someone's personal journal or diary, but it also has some protections from illnesses like depression. First, ALJI searches a journal entry for indicators about the author's health, then ALJI asks the author a few questions to better understand the author, and finally ALJI gives that author information and guidance on improving their health. We are starting to create a computer program of ALJI by first building and testing the detector for the author's health. Instead of making the detector directly, we show the computer some examples of the health indicators from journals we know very well, and then let the computer focus on finding the pattern that would reveal those health indicators from any journal. This is called machine learning, and in our case, ALJI's machine learning is going to be difficult because we have very few example journals where we know all of the health indicators. However, we believe that fixing this issue would solve the first step of ALJI. The end of this thesis also discusses the next steps going forward with ALJI.
38
Vitt, Artur. "Machine Learning in DigitalTelerehabilitation : Telerehabilitation system based on kinect." Thesis, Linnéuniversitetet, Institutionen för datavetenskap (DV), 2015. http://urn.kb.se/resolve?urn=urn:nbn:se:lnu:diva-70048.
Повний текст джерелаАнотація:
The healthcare as a service is always under pressure and is in great demand. Despiteliving in a developed world with access to cars, trains, busses and other transportationmeans, sometimes accessing healthcare can be troublesome and costly. Thecontinuous technological progress provides new means to provide different kind ofservices, healthcare included. One way of putting technology into good use in fieldof healthcare is remote rehabilitation.Remote rehabilitation is a matter of delivering physiotherapy on a distance. Theuse of remote rehabilitation potentially reduces waiting time for treatment and gives apossibility for people with long traveling distance, to be treated at their locations. Thethesis addresses a solution to physiotherapy on distance that utilizes Kinect and machinelearning technologies to provide physiotherapy offline. Thesis presents KinectDigital Rehabilitation Assistant (KiDiRA), which provides simple functions to sufficethe needs of a physiotherapist to plan therapeutical treatment and the ability of apatient to get access physiotherapy offline in real-time at home.More precisely KiDiRA is the system that combines Kinect motion capture device,an interactive graphical interface and a platform to assist with the design ofphysiotherapeutical exercises and an aid for the patient to execute therapeutic plan onhis/her own. The system displays the exercise directives and monitors performanceof patient. KiDiRA aims to incorporate science of machine-learning in process ofperformance evaluation during exercises.
39
Agarwal, Ankur. "Machine Learning for Image Based Motion Capture." Phd thesis, Grenoble INPG, 2006. http://tel.archives-ouvertes.fr/tel-00390301.
Повний текст джерелаАнотація:
Image based motion capture is a problem that has recently gained a lot of attention in the domain of understanding human motion in computer vision. The problem involves estimating the 3D configurations of a human body from a set of images and has applications that include human computer interaction, smart surveillance, video analysis and animation. This thesis takes a machine learning based approach to reconstructing 3D pose and motion from monocular images or video. It makes use of a collection of images and motion capture data to derive mathematical models that allow the recovery of full body configurations directly from image features. The approach is completely data-driven and avoids the use of a human body model. This makes the inference extremely fast. We formulate a class of regression based methods to distill a large training database of motion capture and image data into a compact model that generalizes to predicting pose from new images. The methods rely on using appropriately developed robust image descriptors, learning dynamical models of human motion, and kernelizing the input within a sparse regression framework. Firstly, it is shown how pose can effectively and efficiently be recovered from image silhouettes that are extracted using background subtraction. We exploit sparseness properties of the relevance vector machine for improved generalization and efficiency, and make use of a mixture of regressors for probabilistically handling ambiguities that are present in monocular silhouette based 3D reconstruction. The methods developed enable pose reconstruction from single images as well as tracking motion in video sequences. Secondly, the framework is extended to recover 3D pose from cluttered images by introducing a suitable image encoding that is resistant to changes in background. We show that non-negative matrix factorization can be used to suppress background features and allow the regression to selectively cue on features from the foreground human body. Finally, we study image encoding methods in a broader context and present a novel multi-level image encoding framework called ‘hyperfeatures' that proves to be effective for object recognition and image classification tasks.
40
Letard, Vincent. "Apprentissage incrémental de modèles de domaines par interaction dialogique." Thesis, Université Paris-Saclay (ComUE), 2017. http://www.theses.fr/2017SACLS100/document.
Повний текст джерелаАнотація:
L'intelligence artificielle est la discipline de recherche d'imitation ou de remplacement de fonctions cognitives humaines. À ce titre, l'une de ses branches s'inscrit dans l'automatisation progressive du processus de programmation. Il s'agit alors de transférer de l'intelligence ou, à défaut de définition, de transférer de la charge cognitive depuis l'humain vers le système, qu'il soit autonome ou guidé par l'utilisateur. Dans le cadre de cette thèse, nous considérons les conditions de l'évolution depuis un système guidé par son utilisateur vers un système autonome, en nous appuyant sur une autre branche de l'intelligence artificielle : l'apprentissage artificiel. Notre cadre applicatif est celui de la conception d'un assistant opérationnel incrémental, c'est-à-dire d'un système capable de réagir à des requêtes formulées par l'utilisateur en adoptant les actions appropriées, et capable d'apprendre à le faire. Pour nos travaux, les requêtes sont exprimées en français, et les actions sont désignées par les commandes correspondantes dans un langage de programmation (ici, R ou bash). L'apprentissage du système est effectué à l'aide d'un ensemble d'exemples constitué par les utilisateurs eux-mêmes lors de leurs interactions. Ce sont donc ces derniers qui définissent, progressivement, les actions qui sont appropriées pour chaque requête, afin de rendre le système de plus en plus autonome. Nous avons collecté plusieurs ensembles d'exemples pour l'évaluation des méthodes d'apprentissage, en analysant et réduisant progressivement les biais induits. Le protocole que nous proposons est fondé sur l'amorçage incrémental des connaissances du système à partir d'un ensemble vide ou très restreint. Cela présente l'avantage de constituer une base de connaissances très représentative des besoins des utilisateurs, mais aussi l'inconvénient de n'aquérir qu'un nombre très limité d'exemples. Nous utilisons donc, après examen des performances d'une méthode naïve, une méthode de raisonnement à partir de cas : le raisonnement par analogie formelle. Nous montrons que cette méthode permet une précision très élevée dans les réponses du système, mais également une couverture relativement faible. L'extension de la base d'exemples par analogie est explorée afin d'augmenter la couverture des réponses données. Dans une autre perspective, nous explorons également la piste de rendre l'analogie plus tolérante au bruit et aux faibles différences en entrée en autorisant les approximations, ce qui a également pour effet la production de réponses incorrectes plus nombreuses. La durée d'exécution de l'approche par analogie, déjà de l'ordre de la seconde, souffre beaucoup de l'extension de la base et de l'approximation. Nous avons exploré plusieurs méthodes de segmentation des séquences en entrée afin de réduire cette durée, mais elle reste encore le principal obstacle à contourner pour l'utilisation de l'analogie formelle dans le traitement automatique de la langue. Enfin, l'assistant opérationnel incrémental fondé sur le raisonnement analogique a été testé en condition incrémentale simulée, afin d'étudier la progression de l'apprentissage du système au cours du temps. On en retient que le modèle permet d'atteindre un taux de réponse stable après une dizaine d'exemples vus en moyenne pour chaque type de commande. Bien que la performance effective varie selon le nombre total de commandes considérées, cette propriété ouvre sur des applications intéressantes dans le cadre incrémental du transfert depuis un domaine riche (la langue naturelle) vers un domaine moins riche (le langage de programmation)<br>Artificial Intelligence is the field of research aiming at mimicking or replacing human cognitive abilities. As such, one of its subfields is focused on the progressive automation of the programming process. In other words, the goal is to transfer cognitive load from the human to the system, whether it be autonomous or guided by the user. In this thesis, we investigate the conditions for making a user-guided system autonomous using another subfield of Artificial Intelligence : Machine Learning. As an implementation framework, we chose the design of an incremental operational assistant, that is a system able to react to natural language requests from the user with relevant actions. The system must also be able to learn the correct reactions, incrementally. In our work, the requests are in written French, and the associated actions are represented by corresponding instructions in a programming language (here R and bash). The learning is performed using a set of examples composed by the users themselves while interacting. Thus they progressively define the most relevant actions for each request, making the system more autonomous. We collected several example sets for evaluation of the learning methods, analyzing and reducing the inherent collection biases. The proposed protocol is based on incremental bootstrapping of the system, starting from an empty or limited knowledge base. As a result of this choice, the obtained knowledge base reflects the user needs, the downside being that the overall number of examples is limited. To avoid this problem, after assessing a baseline method, we apply a case base reasoning approach to the request to command transfer problem: formal analogical reasoning. We show that this method yields answers with a very high precision, but also a relatively low coverage. We explore the analogical extension of the example base in order to increase the coverage of the provided answers. We also assess the relaxation of analogical constraints for an increased tolerance of analogical reasoning to noise in the examples. The running delay of the simple analogical approach is already around 1 second, and is badly influenced by both the automatic extension of the base and the relaxation of the constraints. We explored several segmentation strategies on the input examples in order to reduce reduce this time. The delay however remains the main obstacle to using analogical reasoning for natural language processing with usual volumes of data. Finally, the incremental operational assistant based on analogical reasoning was tested in simulated incremental condition in order to assess the learning behavior over time. The system reaches a stable correct answer rate after a dozen examples given in average for each command type. Although the effective performance depends on the total number of accounted commands, this observation opens interesting applicative tracks for the considered task of transferring from a rich source domain (natural language) to a less rich target domain (programming language)
41
Stoffel, Florian [Verfasser]. "Transparency in Interactive Feature-based Machine Learning : Challenges and Solutions / Florian Stoffel." Konstanz : Bibliothek der Universität Konstanz, 2018. http://d-nb.info/1173616314/34.
Повний текст джерела
42
Dondelinger, Frank. "Machine learning approach to reconstructing signalling pathways and interaction networks in biology." Thesis, University of Edinburgh, 2013. http://hdl.handle.net/1842/7850.
Повний текст джерелаАнотація:
In this doctoral thesis, I present my research into applying machine learning techniques for reconstructing species interaction networks in ecology, reconstructing molecular signalling pathways and gene regulatory networks in systems biology, and inferring parameters in ordinary differential equation (ODE) models of signalling pathways. Together, the methods I have developed for these applications demonstrate the usefulness of machine learning for reconstructing networks and inferring network parameters from data. The thesis consists of three parts. The first part is a detailed comparison of applying static Bayesian networks, relevance vector machines, and linear regression with L1 regularisation (LASSO) to the problem of reconstructing species interaction networks from species absence/presence data in ecology (Faisal et al., 2010). I describe how I generated data from a stochastic population model to test the different methods and how the simulation study led us to introduce spatial autocorrelation as an important covariate. I also show how we used the results of the simulation study to apply the methods to presence/absence data of bird species from the European Bird Atlas. The second part of the thesis describes a time-varying, non-homogeneous dynamic Bayesian network model for reconstructing signalling pathways and gene regulatory networks, based on L`ebre et al. (2010). I show how my work has extended this model to incorporate different types of hierarchical Bayesian information sharing priors and different coupling strategies among nodes in the network. The introduction of these priors reduces the inference uncertainty by putting a penalty on the number of structure changes among network segments separated by inferred changepoints (Dondelinger et al., 2010; Husmeier et al., 2010; Dondelinger et al., 2012b). Using both synthetic and real data, I demonstrate that using information sharing priors leads to a better reconstruction accuracy of the underlying gene regulatory networks, and I compare the different priors and coupling strategies. I show the results of applying the model to gene expression datasets from Drosophila melanogaster and Arabidopsis thaliana, as well as to a synthetic biology gene expression dataset from Saccharomyces cerevisiae. In each case, the underlying network is time-varying; for Drosophila melanogaster, as a consequence of measuring gene expression during different developmental stages; for Arabidopsis thaliana, as a consequence of measuring gene expression for circadian clock genes under different conditions; and for the synthetic biology dataset, as a consequence of changing the growth environment. I show that in addition to inferring sensible network structures, the model also successfully predicts the locations of changepoints. The third and final part of this thesis is concerned with parameter inference in ODE models of biological systems. This problem is of interest to systems biology researchers, as kinetic reaction parameters can often not be measured, or can only be estimated imprecisely from experimental data. Due to the cost of numerically solving the ODE system after each parameter adaptation, this is a computationally challenging problem. Gradient matching techniques circumvent this problem by directly fitting the derivatives of the ODE to the slope of an interpolant. I present an inference procedure for a model using nonparametric Bayesian statistics with Gaussian processes, based on Calderhead et al. (2008). I show that the new inference procedure improves on the original formulation in Calderhead et al. (2008) and I present the result of applying it to ODE models of predator-prey interactions, a circadian clock gene, a signal transduction pathway, and the JAK/STAT pathway.
43
Cai, Bingjing. "Machine learning approaches for extracting protein complexes from protein-protein interaction networks." Thesis, Ulster University, 2013. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.603536.
Повний текст джерелаАнотація:
Recent advances in molecular biology have led to the accumulation of large amounts of data on Protein-Protein Interaction (PPI) networks in different species, such as yeast and humans. Due to the inherent complexity, analysing such volumes of data to extract knowledge, such as protein complexes or regulatory pathways, represents not only an enormous challenge but also a great opportunity. This Thesis explores the application of machine learning approaches to detecting protein complexes from PPI networks obtained by Tandem Affinity Purification/Mass Spectrometry (TAP-MS) experiments. TAP-MS PPI networks are usually constructed as binary, and the co-complex relations are largely ignored. In order to take into account the non-binary information of co-complex relations in T AP-MS PPI networks, a new framework for detecting protein complexes has been proposed. Under this framework, two types of graph clustering algorithms and an integrative evaluation platform combining data-driven and knowledge-based quality measures have been proposed and studied. One type of the proposed graph clustering algorithms is random walk based graph clustering, resulting in Enhanced Random Walk with Restart (ERWR) and Random Walk with Restarting Baits (RWRB). The other type is based on the modelling of TAP-MS PPI networks as bipartite graphs, resulting in the Bipartite Graph based Clustering Algorithm (BGCA). The ER WR algorithm has been developed from the Random Walk with Restart (R WR). The key contribution of the ERWR is the introduction of a tuning factor into the random walk process. The tuning factor strengthens connections between nodes that are closer and weakens those that are distant, so that the random walker prefers moving to nodes which are potentially in the same clusters with the starting node.
44
Sarigul, Erol. "Interactive Machine Learning for Refinement and Analysis of Segmented CT/MRI Images." Diss., Virginia Tech, 2004. http://hdl.handle.net/10919/25954.
Повний текст джерелаАнотація:
This dissertation concerns the development of an interactive machine learning method for refinement and analysis of segmented computed tomography (CT) images. This method uses higher-level domain-dependent knowledge to improve initial image segmentation results.
A knowledge-based refinement and analysis system requires the formulation of domain knowledge. A serious problem faced by knowledge-based system designers is the knowledge acquisition bottleneck. Knowledge acquisition is very challenging and an active research topic in the field of machine learning and artificial intelligence. Commonly, a knowledge engineer needs to have a domain expert to formulate acquired knowledge for use in an expert system. That process is rather tedious and error-prone. The domain expert's verbal description can be inaccurate or incomplete, and the knowledge engineer may not correctly interpret the expert's intent. In many cases, the domain experts prefer to do actions instead of explaining their expertise.
These problems motivate us to find another solution to make the knowledge acquisition process less challenging. Instead of trying to acquire expertise from a domain expert verbally, we can ask him/her to show expertise through actions that can be observed by the system. If the system can learn from those actions, this approach is called learning by demonstration.
We have developed a system that can learn region refinement rules automatically. The system observes the steps taken as a human user interactively edits a processed image, and then infers rules from those actions. During the system's learn mode, the user views labeled images and makes refinements through the use of a keyboard and mouse. As the user manipulates the images, the system stores information related to those manual operations, and develops internal rules that can be used later for automatic postprocessing of other images. After one or more training sessions, the user places the system into its run mode. The system then accepts new images, and uses its rule set to apply postprocessing operations automatically in a manner that is modeled after those learned from the human user. At any time, the user can return to learn mode to introduce new training information, and this will be used by the system to updates its internal rule set.
The system does not simply memorize a particular sequence of postprocessing steps during a training session, but instead generalizes from the image data and from the actions of the human user so that new CT images can be refined appropriately.
Experimental results have shown that IntelliPost improves the segmentation accuracy of the overall system by applying postprocessing rules. In tests two different CT datasets of hardwood logs, the use of IntelliPost resulted in improvements of 1.92% and 9.45%, respectively. For two different medical datasets, the use of IntelliPost resulted in improvements of 4.22% and 0.33%, respectively.<br>Ph. D.
45
Iqbal, Sumaiya. "Machine Learning based Protein Sequence to (un)Structure Mapping and Interaction Prediction." ScholarWorks@UNO, 2017. http://scholarworks.uno.edu/td/2379.
Повний текст джерелаАнотація:
Proteins are the fundamental macromolecules within a cell that carry out most of the biological functions. The computational study of protein structure and its functions, using machine learning and data analytics, is elemental in advancing the life-science research due to the fast-growing biological data and the extensive complexities involved in their analyses towards discovering meaningful insights. Mapping of protein’s primary sequence is not only limited to its structure, we extend that to its disordered component known as Intrinsically Disordered Proteins or Regions in proteins (IDPs/IDRs), and hence the involved dynamics, which help us explain complex interaction within a cell that is otherwise obscured. The objective of this dissertation is to develop machine learning based effective tools to predict disordered protein, its properties and dynamics, and interaction paradigm by systematically mining and analyzing large-scale biological data.
In this dissertation, we propose a robust framework to predict disordered proteins given only sequence information, using an optimized SVM with RBF kernel. Through appropriate reasoning, we highlight the structure-like behavior of IDPs in disease-associated complexes. Further, we develop a fast and effective predictor of Accessible Surface Area (ASA) of protein residues, a useful structural property that defines protein’s exposure to partners, using regularized regression with 3rd-degree polynomial kernel function and genetic algorithm. As a key outcome of this research, we then introduce a novel method to extract position specific energy (PSEE) of protein residues by modeling the pairwise thermodynamic interactions and hydrophobic effect. PSEE is found to be an effective feature in identifying the enthalpy-gain of the folded state of a protein and otherwise the neutral state of the unstructured proteins. Moreover, we study the peptide-protein transient interactions that involve the induced folding of short peptides through disorder-to-order conformational changes to bind to an appropriate partner. A suite of predictors is developed to identify the residue-patterns of Peptide-Recognition Domains from protein sequence that can recognize and bind to the peptide-motifs and phospho-peptides with post-translational-modifications (PTMs) of amino acid, responsible for critical human diseases, using the stacked generalization ensemble technique. The involved biologically relevant case-studies demonstrate possibilities of discovering new knowledge using the developed tools.
46
Wang, Jianxiong. "A machine learning system for understanding appraisal in design documents." Thesis, The University of Sydney, 2009. https://hdl.handle.net/2123/28237.
Повний текст джерелаАнотація:
This thesis describes a machine learning system for understanding appraisal in design
documents. In linguistics, analysis of attitudes expressed in text is known as appraisal,
which is the writer's subjective attitude (i.e., the opinion, sentiment or attitude expressed)
toward the semantic meaning. Semantic orientation is the aspect of appraisal that
expresses the positive or negative stance of the author to the semantic meaning.
Computational linguistics methods have been introduced into the area of understanding
appraisal. In computational linguistics, the analysis of appraisal is known as sentiment
analysis.
One of the identified research issues in sentiment analysis is the lack of a precise and
agreed-upon computational language model of sentiment. At issue is whether the
incorporation of knowledge about the grammatical structure of appraisals would improve
automated semantic orientation classification over so-called “bag-of—words” approaches.
Another research issue is the extent to which the portability of a sentiment analysis
system is limited by the training data set. This research takes a different approach by
using machine learning to infer which word group indicators are useful for sentiment
analysis in a compact representation of text.
To address these research issues, this thesis aims to construct a sentiment analysis system
based on a compact 9-dimensional representation of text that is lexicon-independent. This
research takes a statistical supervised machine learning approach to understand semantic
meaning and sentiment specifically in design documents. The fundamental thrust of this
research program is two-fold: first, to create a computational language model to represent
the semantic orientation of appraisal and semantic meaning in design documents; and
secondly, to implement and test the model.
47
Lynch, Paul Kieran. "The generation of knowledge based systems for interactive nonlinear constrained optimisation." Thesis, Queen's University Belfast, 1997. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.388221.
Повний текст джерела
48
Sanchez, Téo. "Interactive Machine Teaching with and for Novices." Electronic Thesis or Diss., université Paris-Saclay, 2022. http://www.theses.fr/2022UPASG055.
Повний текст джерелаАнотація:
Les algorithmes d'apprentissage machine déployés dans la société ou la technologie offrent généralement aux utilisateurs aucune prise sur la manière dont les modèles d'apprentissage sont optimisés à partir des données. Seuls les experts conçoivent, analysent et optimisent les algorithmes d'apprentissage automatique. À l'intersection de l'Interaction Humain-Machine (IHM) et de l'apprentissage machine, le domaine de l'apprentissage automatique interactif vise à intégrer l'apprentissage automatique dans des pratiques existantes. L'enseignement machine interactif (Interactive Machine Teaching), en particulier, cherche à impliquer des utilisateurs non experts en tant qu'enseignant de la machine afin de les autonomiser dans le processus de construction de modèles d'apprentissage. Ces utilisateurs pourraient profiter de la construction de modèles d'apprentissage pour traiter et automatiser des tâches sur leurs propres données, conduisant à des modèles plus robustes et moins biaisés pour des problèmes spécialisés. Cette thèse adopte une approche empirique sur l'enseignement machine interactif en se concentrant sur la façon dont les utilisateurs développent des stratégies et comprennent les systèmes d'apprentissage machine interactifs à travers l'acte d'enseigner. Cette recherche fournit deux études utilisateurs impliquant des participants en tant qu'enseignant de classificateurs d'images utilisant des réseaux de neurones artificiels appris par transfert. Ces études se concentrent sur ce que les utilisateurs comprennent du comportement du modèle ML et sur la stratégie qu'ils peuvent utiliser pour le "faire fonctionner". La seconde étude se concentre sur la compréhension et l'utilisation de deux types d'incertitude : l'incertitude aléatorique, qui traduit l'ambiguïté, et l'incertitude épistémique, qui traduit la nouveauté. Je discute de l'utilisation de l'incertitude et de l'apprentissage actif (Active Learning) comme outils pour l'enseignement machine interactif. Enfin, je présente mes collaborations artistiques et adopte une approche réflexive sur les obstacles et les opportunités de développement de l'apprentissage automatique interactif pour l'art. Je soutiens que les utilisateurs novices développent différentes stratégies d'enseignement qui peuvent évoluer en fonction des informations obtenues tout au long de l'interaction. Les stratégies d'enseignement structurent la composition des données d'entraînement et affectent la capacité des utilisateurs à comprendre et à prédire le comportement de l'algorithme. En plus de permettre aux gens de construire des modèles d'apprentissage automatique, l'enseignement machine interactif présente un intérêt pédagogique en favorisant les comportements d'investigation, renforçant les connaissances des novices en apprentissage machine<br>Machine Learning algorithms in society or interactive technology generally provide users with little or no agency with respect to how models are optimized from data. Only experts design, analyze, and optimize ML algorithms. At the intersection of HCI and ML, the field of Interactive Machine Learning (IML) aims at incorporating ML workflows within existing users' practices. Interactive Machine Teaching (IMT), in particular, focuses on involving non-expert users as "machine teachers" and empowering them in the process of building ML models. Non-experts could take advantage of building ML models to process and automate tasks on their data, leading to more robust and less biased models for specialized problems. This thesis takes an empirical approach to IMT by focusing on how people develop strategies and understand interactive ML systems through the act of teaching. This research provides two user studies involving participants as teachers of image-based classifiers using transfer-learned artificial neural networks. These studies focus on what users understand from the ML model's behavior and what strategy they may use to "make it work." The second study focuses on machine teachers' understanding and use of two types of uncertainty: aleatoric uncertainty, which conveys ambiguity, and epistemic uncertainty, which conveys novelty. I discuss the use of uncertainty and active learning in IMT. Finally, I report artistic collaborations and adopt an auto-ethnographic approach to challenges and opportunities for developing IMT with artists. I argue that people develop different teaching strategies that can evolve with insights obtained throughout the interaction. People's teaching strategies structure the composition of the data they curated and affect their ability to understand and predict the algorithm behavior. Besides empowering people to build ML models, IMT can foster investigative behaviors, leveraging peoples' literacy in ML and artificial intelligence
49
Billewicz, Agnieszka. "Study of a relationship. Designerly explorations of machine learning algorithms." Thesis, Malmö högskola, Fakulteten för kultur och samhälle (KS), 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:mau:diva-22593.
Повний текст джерелаАнотація:
Study of a relationship is a 10-week Research by Design project that explores the space of intersection between Design and Machine Learning. It is a series of design engagements and experiments, heavily grounded in the present time and simple technology, that produces a semi-abstract knowledge on relationship that can be established between human and Machine Learning artefacts. This research strives to propose an alternative designerly approach towards Machine Learning, one that would promote evoking positive emotions, usage for personal purposes and understanding of basic principles behind technology, thus putting the human in the position of control.
50
Akrour, Riad. "Robust Preference Learning-based Reinforcement Learning." Thesis, Paris 11, 2014. http://www.theses.fr/2014PA112236/document.
Повний текст джерелаАнотація:
Les contributions de la thèse sont centrées sur la prise de décisions séquentielles et plus spécialement sur l'Apprentissage par Renforcement (AR). Prenant sa source de l'apprentissage statistique au même titre que l'apprentissage supervisé et non-supervisé, l'AR a gagné en popularité ces deux dernières décennies en raisons de percées aussi bien applicatives que théoriques. L'AR suppose que l'agent (apprenant) ainsi que son environnement suivent un processus de décision stochastique Markovien sur un espace d'états et d'actions. Le processus est dit de décision parce que l'agent est appelé à choisir à chaque pas de temps du processus l'action à prendre. Il est dit stochastique parce que le choix d'une action donnée en un état donné n'implique pas le passage systématique à un état particulier mais définit plutôt une distribution sur l'espace d'états. Il est dit Markovien parce que cette distribution ne dépend que de l'état et de l'action courante. En conséquence d'un choix d'action, l'agent reçoit une récompense. Le but de l'AR est alors de résoudre le problème d'optimisation retournant le comportement qui assure à l'agent une récompense maximale tout au long de son interaction avec l'environnement. D'un point de vue pratique, un large éventail de problèmes peuvent être transformés en un problème d'AR, du Backgammon (cf. TD-Gammon, l'une des premières grandes réussites de l'AR et de l'apprentissage statistique en général, donnant lieu à un joueur expert de classe internationale) à des problèmes de décision dans le monde industriel ou médical. Seulement, le problème d'optimisation résolu par l'AR dépend de la définition préalable d'une fonction de récompense adéquate nécessitant une expertise certaine du domaine d'intérêt mais aussi du fonctionnement interne des algorithmes d'AR. En ce sens, la première contribution de la thèse a été de proposer un nouveau cadre d'apprentissage, allégeant les prérequis exigés à l'utilisateur. Ainsi, ce dernier n'a plus besoin de connaître la solution exacte du problème mais seulement de pouvoir désigner entre deux comportements, celui qui s'approche le plus de la solution. L'apprentissage se déroule en interaction entre l'utilisateur et l'agent. Cette interaction s'articule autour des trois points suivants : i) L'agent exhibe un nouveau comportement ii) l'expert le compare au meilleur comportement jusqu'à présent iii) l'agent utilise ce retour pour mettre à jour son modèle des préférences puis choisit le prochain comportement à démontrer. Afin de réduire le nombre d'interactions nécessaires entre l'utilisateur et l'agent pour que ce dernier trouve le comportement optimal, la seconde contribution de la thèse a été de définir un critère théoriquement justifié faisant le compromis entre les désirs parfois contradictoires de prendre en compte les préférences de l'utilisateur tout en exhibant des comportements suffisamment différents de ceux déjà proposés. La dernière contribution de la thèse est d'assurer la robustesse de l'algorithme face aux éventuelles erreurs d'appréciation de l'utilisateur. Ce qui arrive souvent en pratique, spécialement au début de l'interaction, quand tous les comportements proposés par l'agent sont loin de la solution attendue<br>The thesis contributions resolves around sequential decision taking and more precisely Reinforcement Learning (RL). Taking its root in Machine Learning in the same way as supervised and unsupervised learning, RL quickly grow in popularity within the last two decades due to a handful of achievements on both the theoretical and applicative front. RL supposes that the learning agent and its environment follow a stochastic Markovian decision process over a state and action space. The process is said of decision as the agent is asked to choose at each time step an action to take. It is said stochastic as the effect of selecting a given action in a given state does not systematically yield the same state but rather defines a distribution over the state space. It is said to be Markovian as this distribution only depends on the current state-action pair. Consequently to the choice of an action, the agent receives a reward. The RL goal is then to solve the underlying optimization problem of finding the behaviour that maximizes the sum of rewards all along the interaction of the agent with its environment. From an applicative point of view, a large spectrum of problems can be cast onto an RL one, from Backgammon (TD-Gammon, was one of Machine Learning first success giving rise to a world class player of advanced level) to decision problems in the industrial and medical world. However, the optimization problem solved by RL depends on the prevous definition of a reward function that requires a certain level of domain expertise and also knowledge of the internal quirks of RL algorithms. As such, the first contribution of the thesis was to propose a learning framework that lightens the requirements made to the user. The latter does not need anymore to know the exact solution of the problem but to only be able to choose between two behaviours exhibited by the agent, the one that matches more closely the solution. Learning is interactive between the agent and the user and resolves around the three main following points: i) The agent demonstrates a behaviour ii) The user compares it w.r.t. to the current best one iii) The agent uses this feedback to update its preference model of the user and uses it to find the next behaviour to demonstrate. To reduce the number of required interactions before finding the optimal behaviour, the second contribution of the thesis was to define a theoretically sound criterion making the trade-off between the sometimes contradicting desires of complying with the user's preferences and demonstrating sufficiently different behaviours. The last contribution was to ensure the robustness of the algorithm w.r.t. the feedback errors that the user might make. Which happens more often than not in practice, especially at the initial phase of the interaction, when all the behaviours are far from the expected solution