To see the other types of publications on this topic, follow the link: Combined Learning.
Dissertations / Theses on the topic 'Combined Learning'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Combined Learning.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
1
Paulus, Kyle. "Combined CALPHAD and Machine Learning for Property Modelling." Thesis, KTH, Materialvetenskap, 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-278149.
Full textAbstract:
Techniques to improve the speed at which materials are researched and developed has been conducted by investigating the machine learning methodology. These techniques offer solutions to connect the length scales of material prop- erties from atomistic and chemical features using materials databases generated from collected data. In this assessment, two material informatics methodologies are used to predict material properties in steels and nickel based superalloys using this approach. Martensite start temperature and sigma phase amount as a function of input composition has been modelled with the use of machine learning algorithms. The experimental methodology had a collection of over 2000 unique experimental martensite start temperature points. This yielded important information on higher order interactions for the martensite start temperature, and a root mean square error (rmse) of 29 Kelvin using ensemble tree based algorithms. The metamodel was designed using an artificial neural network from TensorFlow’s library to predict sigma phase fraction and its composition. The methodology for building, calculating, and using data from TC-Python will be laid out. This generates a model that would generalize sigma phase fraction 97.9 % of Thermo-Calc’s equilibrium model in 7.1 seconds compared to 227 hours neded in the simulation to calculate the same amount of material property data.<br>Tekniker för att förbättra hastigheten med material som forskas och utvecklas har genomförts genom att undersöka metodik för maskininlärning. Dessa tekniker erbjuder lösningar för att ansluta längdskalorna för materialegenskaper från atomistiska och kemiska egenskaper med hjälp av materialdatabaser genererade från insamlade data. I denna bedömning används två materialinformatikmetoder för att förutsäga materialegenskaper i stål och nickelbaserade superlegeringar med denna metod. Martensite-starttemperatur och sigmafasmängd som en funktion av ingångssammansättningen har modellerats med användning av maskininlärningsalgoritmer. Den experimentella metoden hade en samling av över 2000 unika experimentella starttemperaturpunkter för martensit. Detta gav viktig information om interaktioner med högre ordning för martensit-starttemperaturen och ett root-medelvärde-kvadratfel (rmse) på 29 Kelvin med användning av ensemble-trädbaserade algoritmer. Metamodellen designades med hjälp av ett artificiellt neuralt nätverk från TensorFlows bibliotek för att förutsäga sigma-fasfraktion och dess sammansättning. Metoden för att bygga, beräkna och använda data från TC-Python kommer att anges. Detta genererar en modell som skulle generalisera sigma-fasfraktion 97,9 % av Thermo-Calcs jämviktsmodell på 7,1 sekunder jämfört med 227 timmar som behövs i simuleringen för att beräkna samma mängd materialegenskapsdata.
2
Olsen, Jennifer. "Orchestrating Combined Collaborative and Individual Learning in the Classroom." Research Showcase @ CMU, 2017. http://repository.cmu.edu/dissertations/1000.
Full textAbstract:
In the classroom, teachers make use of different combinations of social planes (e.g., individual, collaborative) to support learning. However, little is known about the complementary strengths of individual and collaborative learning or how to combine them so that they are more effective than either social plane alone. One roadblock to this investigation is an ability to orchestrate, or manage, more complex, but theoretically effective, combinations of collaborative and individual learning in the classroom. Prior research has created orchestration tools that support the planning and real-time management of classroom activities, which reduces the cognitive load and time needed for instructors to support the activity, allowing for more complex activities to become more manageable. Current orchestration tools do not, however, support a wide range of combinations of collaborative and individual learning activities in a flexible manner. To fully investigate the combinations of collaborative and individual learning, orchestration tools need to be developed that can support the researcher in a way that can be integrated into the classroom by accounting for teachers’ values. My thesis work addresses two related goals. First, my work addresses the questions: Do collaborative and individual learning have complementary strengths and is a combination of the two social planes better than either alone? In my work, I developed an intelligent tutoring system (ITS) to support collaborative and individual learning. Through three studies, using this ITS, with over 500 4th and 5th grade students, I demonstrate that a collaborative ITS can be used to effectively support learning with elementary school students and that a combination of collaborative and individual learning is more effective than either alone. However, my studies did not find any support for complementary strengths and many other combinations of social planes are left to investigate. Additionally, during my experiments, I encountered challenges in orchestration that, along with the need to research more complex combinations of collaborative and individual learning, informed the next steps of my research. The second question my thesis work addresses is: How does an orchestration tool that supports researchers in exploring this space need to be designed to align with teachers’ values for easy integration in the classroom? Specifically, I aimed to support fluid transitions between social planes where students do not all have to be working in sync, which is not currently supported in existing orchestration tools. To support the orchestration tool design, I present a framework that structures the space that a researcher can explore when combining individual and collaborative learning. The framework can act as the set of requirements to be met in the orchestration tool from the point of the researcher as well as a lens to analyze and design combined social plane activities. As a first step towards supporting fluid transitions as laid out in the framework, I present a set of statistical models that extend domain-level individual modeling into the space of collaborative environments. Finally, I developed an orchestration prototype built around my framework that can be used as a research tool to further explore combined collaborative and individual spaces. To develop the tool to be successful within the classroom, I worked with teachers through a co-design process and validation of the prototype to incorporate their values into the tool. Taken together, my dissertation has six primary contributions. My dissertation contributes to the learning sciences through advancing our knowledge of (1) the strengths of collaborative and individual learning, although I did not find any complementary strengths, and (2) if a combination is better than either alone, which I did find support for. It contributes to educational technology through (3) the design of an effective ITS that supports collaborative and individual learning for fractions and educational data mining through (4) the advancement of models that can more accurately predict individual learning within a collaborative setting than the existing individual models. Finally, it contributes to computer supported collaborative learning and human-computer interaction through (5) a framework, which provides a lens for designing and analyzing combined collaborative and individual learning spaces, and (6) an orchestration prototype that supports fluid transitions between social planes in a way that can be a useful to both researchers and teachers in the classroom.
3
Ko, E. Soon. "Product Matching through Multimodal Image and Text Combined Similarity Matching." Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-301306.
Full textAbstract:
Product matching in e-commerce is an area that faces more and more challenges with growth in the e-commerce marketplace as well as variation in the quality of data available online for each product. Product matching for e-commerce provides competitive possibilities for vendors and flexibility for customers by identifying identical products from different sources. Traditional methods in product matching are often conducted through rule-based methods and methods tackling the issue through machine learning usually do so through unimodal systems. Moreover, existing methods would tackle the issue through product identifiers which are not always unified for each product. This thesis provides multimodal approaches through product name, description, and image to the problem area of product matching that outperforms unimodal approaches. Three multimodal approaches were taken, one unsupervised and two supervised. The unsupervised approach uses straight-forward embedding space to nearest neighbor search that provides better results than unimodal approaches. One of the supervised multimodal approaches uses Siamese network on the embedding space which outperforms the unsupervised multi- modal approach. Finally, the last supervised approach instead tackles the issue by exploiting distance differences in each modality through logistic regression and a decision system that provided the best results.<br>Produktmatchning inom e-handel är ett område som möter fler och fler utmaningar med hänsyn till den tillväxt som e-handelsmarknaden undergått och fortfarande undergår samt variation i kvaliteten på den data som finns tillgänglig online för varje produkt. Produktmatchning inom e-handel är ett område som ger konkurrenskraftiga möjligheter för leverantörer och flexibilitet för kunder genom att identifiera identiska produkter från olika källor. Traditionella metoder för produktmatchning genomfördes oftast genom regelbaserade metoder och metoder som utnyttjar maskininlärning gör det vanligtvis genom unimodala system. Dessutom utnyttjar mestadels av befintliga metoder produktidentifierare som inte alltid är enhetliga för varje produkt mellan olika källor. Denna studie ger istället förslag till multimodala tillvägagångssätt som istället använder sig av produktnamn, produktbeskrivning och produktbild för produktmatchnings-problem vilket ger bättre resultat än unimodala metoder. Tre multimodala tillvägagångssätt togs, en unsupervised och två supervised. Den unsupervised metoden använder embeddings vektorerna rakt av för att göra en nearest neighborsökning vilket gav bättre resultat än unimodala tillvägagångssätt. Ena supervised multimodal tillvägagångssätten använder siamesiska nätverk på embedding utrymmet vilket gav resultat som överträffade den unsupervised multimodala tillvägagångssättet. Slutligen tar den sista supervised metoden istället avståndsskillnader i varje modalitet genom logistisk regression och ett beslutssystem som gav bästa resultaten.
4
Bhatt, Dhruv. "Economic Dispatch of the Combined Cycle Power Plant Using Machine Learning." Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2019. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-266110.
Full textAbstract:
Combined Cycle Power Plant (CCPP)s play a key role in modern powersystem due to their lesser investment cost, lower project executiontime, and higher operational flexibility compared to other conventionalgenerating assets. The nature of generation system is changing withever increasing penetration of the renewable energy resources. Whatwas once a clearly defined generation, transmission, and distributionflow is shifting towards fluctuating distribution generation. Because ofvariation in energy production from the renewable energy resources,CCPP are increasingly required to vary their load levels to keep balancebetween supply and demand within the system. CCPP are facingmore number of start cycles. This induces more stress on the gas turbineand as a result, maintenance intervals are affected.The aim of this master thesis project is to develop a dispatch algorithmfor the short-term operation planning for a combined cyclepower plant which also includes the long-term constraints. The longtermconstraints govern the maintenance interval of the gas turbines.These long-term constraints are defined over number of EquivalentOperating Hours (EOH) and Equivalent Operating Cycles (EOC) forthe Gas Turbine (GT) under consideration. CCPP is operating in theopen electricity market. It consists of two SGT-800 GT and one SST-600 Steam Turbine (ST). The primary goal of this thesis is to maximizethe overall profit of CCPP under consideration. The secondary goal ofthis thesis it to develop the meta models to estimate consumed EOHand EOC during the planning period.Siemens Industrial Turbo-machinery AB (SIT AB) has installed sensorsthat collects the data from the GT. Machine learning techniqueshave been applied to sensor data from the plant to construct Input-Output (I/O) curves to estimate heat input and exhaust heat. Resultsshow potential saving in the fuel consumption for the limit on CumulativeEquivalent Operating Hours (CEOH) and Cumulative EquivalentOperating Cycles (CEOC) for the planning period. However, italso highlighted some crucial areas of improvement before this economicdispatch algorithm can be commercialized.<br>Kombicykelkraftverk spelar en nyckelroll i det moderna elsystemet pågrund av den låga investeringskostnaden, den korta tiden för att byggaett nytta kraftverk och hög flexibilitet jämfört med andra kraftverk.Elproduktionssystemen förändras i takt med en allt större andel förnybarelproduktion. Det som en gång var ett tydligt definierat flödefrån produktion via transmission till distribution ändrar nu karaktärtill fluktuerande, distribuerad generering. På grund av variationernai elproduktion från förnybara energikällor finns ett ökat behov avatt kombicykelkraftverk varierar sin elproduktion för att upprätthållabalansen mellan produktion och konsumtion i systemet. Kombicykelkraftverkbehöver startas och stoppas oftare. Detta medför mer stresspå gasturbinen och som ett resultat påverkas underhållsintervallerna.Syftet med detta examensarbete är att utveckla en algoritm för korttidsplaneringav ett kombicykelkraftverk där även driften på lång siktbeaktas. Begränsningarna på lång sikt utgår från underhållsintervallenför gasturbinerna. Dessa långsiktiga begränsningar definieras som antaletekvivalenta drifttimmar och ekvivalenta driftcykler för det aktuellakraftverket. Kombikraftverket drivs på den öppna elmarknaden.Det består av två SGT-800 GT och en SST-600 ångturbin. Det främstamålet med examensarbetet är att maximera den totala vinsten förkraftverket. Ett sekundärt mål är att utveckla metamodeller för attskatta använda ekvivalenta drifttimmar och ekvivalenta driftcyklerunder planeringsperioden.Siemens Industrial Turbo-machinery AB (SIT AB) har installeratsensorer som samlar in data från gasturbinerna. Maskininlärningsteknikerhar tillämpats på sensordata för att konstruera kurvor för attuppskatta värmetillförseln och avgasvärme. Resultaten visar en potentiellbesparing i bränsleförbrukningen om de sammanlagda ekvivalentadrifttimmarna och de sammanlagda ekvivalenta driftcyklernabegränsas under planeringsperioden. Det framhålls dock också att detfinns viktiga förbättringar som behövs innan korttidsplaneringsalgoritmenkan kommersialiseras.
5
Thornton, Chris. "Auto-WEKA : combined selection and hyperparameter optimization of supervised machine learning algorithms." Thesis, University of British Columbia, 2014. http://hdl.handle.net/2429/46177.
Full textAbstract:
Many different machine learning algorithms exist; taking into account each algorithm's set of hyperparameters, there is a staggeringly large number of possible choices. This project considers the problem of simultaneously selecting a learning algorithm and setting its hyperparameters. Previous works attack these issues separately, but this problem can be addressed by a fully automated approach, in particular by leveraging recent innovations in Bayesian optimization. The WEKA software package provides an implementation for a number of feature selection and supervised machine learning algorithms, which we use inside our automated tool, Auto-WEKA. Specifically, we examined the 3 search and 8 evaluator methods for feature selection, as well as all of the classification and regression methods, spanning 2 ensemble methods, 10 meta-methods, 27 base algorithms, and their associated hyperparameters. On 34 popular datasets from the UCI repository, the Delve repository, the KDD Cup 09, variants of the MNIST dataset and CIFAR-10, our method produces classification and regression performance often much better than obtained using state-of-the-art algorithm selection and hyperparameter optimization methods from the literature. Using this integrated approach, users can more effectively identify not only the best machine learning algorithm, but also the corresponding hyperparameter settings and feature selection methods appropriate for that algorithm, and hence achieve improved performance for their specific classification or regression task.
6
Desypris, Georgios A. "Enhancement of learning process in web-based courses using combined media components." Thesis, Monterey, Calif. : Springfield, Va. : Naval Postgraduate School ; Available from National Technical Information Service, 2002. http://library.nps.navy.mil/uhtbin/hyperion-image/02sep%5FDesypris.pdf.
Full textAbstract:
Thesis (M.S. in Computer Science)--Naval Postgraduate School, September 2002.<br>Thesis advisor(s): Rudolph Darken, Anthony Ciavarelli. Includes bibliographical references (p. 85-86). Also available online.
7
Lombard, Valentin. "Geometric deep manifold learning combined with natural language processing for protein movies." Electronic Thesis or Diss., Sorbonne université, 2024. http://www.theses.fr/2024SORUS379.
Full textAbstract:
Les protéines jouent un rôle central dans les processus biologiques, et comprendre comment elles se déforment et se déplacent est essentiel pour élucider leurs mécanismes fonctionnels. Malgré les récentes avancées dans les technologies à haut débit, qui ont élargi nos connaissances sur les structures protéiques, la prédiction précise de leurs différents états conformationnels et mouvements reste un défi majeur. Nous présentons deux approches complémentaires pour relever le défi de la compréhension et de la prédiction de l'ensemble de la variabilité conformationnelle des protéines. La première approche, appelée Dimensionality Analysis for protein Conformational Exploration (DANCE), permet une description systématique et complète de la variabilité conformationnelle des familles de protéines. DANCE prend en compte à la fois les structures expérimentales et prédites. Elle est adaptée à l'analyse des protéines individuelles jusqu'aux superfamilles. En l'utilisant, nous avons regroupé toutes les structures protéiques résolues expérimentalement disponibles dans la banque de données Protein Data Bank en collections conformationnelles et les avons caractérisées comme des ensembles de mouvements linéaires. Cette ressource facilite l'accès et l'exploitation des multiples états adoptés par une protéine et ses homologues. Au-delà de l'analyse descriptive, nous avons évalué des techniques classiques de réduction de la dimensionnalité pour échantillonner des états non observés sur un banc d'essai représentatif. Ce travail améliore notre compréhension de la manière dont les protéines se déforment pour accomplir leurs fonctions et ouvre la voie à une évaluation standardisée des méthodes conçues pour échantillonner et générer des conformations protéiques. La deuxième approche repose sur l'apprentissage profond pour prédire des représentations continues du mouvement des protéines directement à partir de séquences, sans avoir besoin de données structurelles. Ce modèle, appelé SeaMoon, utilise des embeddings de modèles de langage protéique (pLM) comme entrées dans un réseau neuronal convolutif léger comptant environ un million de paramètres entraînables. SeaMoon atteint un taux de réussite de 40 % lorsqu'il est évalué sur environ 1 000 collections de conformations expérimentales, capturant des mouvements au-delà de la portée des méthodes traditionnelles comme l'analyse des modes normaux, qui repose uniquement sur la géométrie 3D. De plus, SeaMoon se généralise à des protéines n'ayant aucune similitude de séquence détectable avec son ensemble d'entraînement et peut être facilement réentraîné avec des pLM mis à jour. Ces deux approches offrent un cadre unifié pour faire progresser notre compréhension de la dynamique des protéines. DANCE fournit une exploration détaillée des mouvements protéiques basée sur des données structurelles, tandis que SeaMoon démontre le potentiel des modèles d'apprentissage profond basés sur les séquences pour capturer des mouvements complexes sans dépendre d'informations structurelles explicites. Ensemble, elles ouvrent la voie à une compréhension plus complète de la variabilité conformationnelle des protéines et de son rôle dans la fonction biologique<br>Proteins play a central role in biological processes, and understanding how they deform and move is essential to elucidating their functional mechanisms. Despite recent advances in high-throughput technologies, which have broadened our knowledge of protein structures, accurate prediction of their various conformational states and motions remains a major challenge. We present two complementary approaches to address the challenge of understanding and predicting the full range of protein conformational variability. The first approach, Dimensionality Analysis for protein Conformational Exploration (DANCE) for a systematic and comprehensive description of protein families conformational variability. DANCE accommodates both experimental and predicted structures. It is suitable for analyzing anything from single proteins to superfamilies. Employing it, we clustered all experimentally resolved protein structures available in the Protein Data Bank into conformational collections and characterized them as sets of linear motions. The resource facilitates access and exploitation of the multiple states adopted by a protein and its homologs. Beyond descriptive analysis, we assessed classical dimensionality reduction techniques for sampling unseen states on a representative benchmark. This work improves our understanding of how proteins deform to perform their functions and opens ways to a standardized evaluation of methods designed to sample and generate protein conformations. The second approach relies on deep learning to predict continuous representations of protein motion directly from sequences, without the need for structural data. This model, SeaMoon, uses protein language model (pLM) embeddings as inputs to a lightweight convolutional neural network with around 1 million trainable parameters. SeaMoon achieves a success rate of 40% when evaluated against around 1,000 collections of experimental conformations, capturing movements beyond the reach of traditional methods such as normal mode analysis, which relies solely on 3D geometry. In addition, SeaMoon generalizes to proteins that have no detectable sequence similarity with its training set and can be easily retrained with updated pLMs. These two approaches offer a unified framework for advancing our understanding of protein dynamics. DANCE provides a detailed exploration of protein movements based on structural data, while SeaMoon demonstrates the potential of sequence-based deep learning models to capture complex movements without relying on explicit structural information. Together, they pave the way for a more comprehensive understanding of protein conformational variability and its role in biological function
8
Agarwal, Akrita. "Exploring the Noise Resilience of Combined Sturges Algorithm." University of Cincinnati / OhioLINK, 2015. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1447070335.
Full text
9
Mazzitelli, Carla. "Aprendizado motor em escolares: comparação entre prática mental, prática física e prática combinada." Universidade de São Paulo, 2013. http://www.teses.usp.br/teses/disponiveis/47/47135/tde-28062013-142756/.
Full textAbstract:
A compreensão do processo de aprendizagem motora é de real importância para a atividade fisioterapêutica, visto que as abordagens e técnicas visam à mudança e o aperfeiçoamento de uma atividade motora ou mesmo a aquisição de uma nova habilidade. Neste processo a prática realizada tem implicação direta na aprendizagem motora. Objetivo: comparar os efeitos da prática combinada (iniciada pelo treino mental, seguido pelo treino físico, com distribuição proporcionada entre os mesmos), com os efeitos da prática mental e física isoladas, sobre aquisição, retenção, transferência inter tarefa e inter manual em crianças de 9 a 10 anos. Método: participaram do estudo 36 crianças, divididas em 3 grupos de acordo com o tipo de prática realizada: o grupo PM está composto 12 crianças, o PF por 12 crianças e o PC por 12 crianças. O treinamento foi realizado em sessão única e constituiu da repetição de 2400 movimentos de uma sequência de movimentos de oposição dos dedos. Todos os grupos foram submetidos às mesmas avaliações antes do treinamento, depois do treinamento, 4, 7, e 28 dias após o treinamento. A aquisição, a retenção e a transferência foram avaliados por meio da análise de variância (ANOVA). Para todas as interações que alcançaram nível de significância (p<0,05) foi aplicado o pós-teste de Tukey. Resultados: para o tempo de treinamento, tanto a PM quanto a PF proporcionaram melhora progressiva da velocidade ao longo dos blocos, já a PC proporcionou melhora inicial entre os dois primeiros blocos, e piora na velocidade do segundo para o terceiro que não se modificou até o final do treino. A análise da aquisição evidenciou que a PC proporcionou ao final do treino o mesmo nível de desempenho da PM, mas inferior ao obtido pela PF. Para retenção, a PC proporcionou os mesmo níveis obtidos pela PF e PM, com ganhos adicionais tardios. Para a transferência inter-sequência e intermanual, a PC permitiu a transferência do desempenho obtido por meio do treino da ST para a SR de forma similar a PM, mas superior a PF. Conclusão: a prática combinada não mostrou vantagens em comparação a prática física ou mental isoladas para a aprendizagem, retenção de uma tarefa de oposição sequencial de dedos em crianças de 9 e 10 anos. Entretanto, mostrou-se mais eficiente que a prática física isolada para a transferência inter-sequências e intermanual, alcançando a mesma eficiência da prática mental isolada. Assim, as três formas de prática foram igualmente eficientes para a aprendizagem e retenção de movimentos sequenciais e a prática mental, combinada ou não com a prática física foi mais eficiente que a prática física isolada para a transferência inter-sequências e intermanual<br>Understanding the process of motor learning it is important to physiotherapy so that approaches and techniques designed to change and improve motor activity or even the acquisition of a new skill. In this process the practice has held direct implication in motor learning. Purpose: To compare the effects of combined practice (initiated by mental training, followed by physical training, with distribution provided between them), with the effects of mental and physical practice isolated on acquisition, retention and transfer on the learning of sequential finger opposition movements in children 9-10 years. Methods: This study included thirty-six 9 to 10 year-old children that were randomly allocated into three groups: mental practice (MP), physical practice (PP), and combined (PC). The training was conducted in a single session and consisted repetition of 2400 movements sequence of finger opposition movements. All groups underwent the same evaluations before training, after training, 4, 7, and 28 days after training. The acquisition, retention and transfer were evaluated by analysis of variance (ANOVA). For all interactions that reached statistical significance (p <0.05) was applied post-Tukey test. Results: The training time for both the MP and FP provided progressive improvement in speed over the blocks, the CP has provided initial improvement between the first two blocks, and worsened speed from the second to the third that did not change until the end of the training. The analysis showed that acquisition of the CP at the end of the training provided the same level of performance of the MP, but lower than that obtained by the FP. For retention, the CP showed the same levels obtained by the FP and MP, with additional gains late. For inter-sequence and intermanual transfer, CP allowed to transfer performance obtained by training for the trained sequence similarly to reverse sequence by MP, but exceeding FP. Conclusion: The combined practice showed no advantages compared to physical or mental practice isolated to learning retention task of sequential finger opposition in children aged 9 and 10 years. However, proved more efficient than single physical practice to transfer inter-sequences and intermanual, achieving the same efficiency of mental practice alone. So the three forms of practice were equally efficient for learning and retention of sequential movements and mental practice, whether or not combined with physical practice was more effective than physical practice alone for transfer
10
Carls, Fredrik. "Evaluation of machine learning methods for anomaly detection in combined heat and power plant." Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2019. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-255006.
Full textAbstract:
In the hope to increase the detection rate of faults in combined heat and power plant boilers thus lowering unplanned maintenance three machine learning models are constructed and evaluated. The algorithms; k-Nearest Neighbor, One-Class Support Vector Machine, and Auto-encoder have a proven track record in research for anomaly detection, but are relatively unexplored for industrial applications such as this one due to the difficulty in collecting non-artificial labeled data in the field.The baseline versions of the k-Nearest Neighbor and Auto-encoder performed very similarly. Nevertheless, the Auto-encoder was slightly better and reached an area under the precision-recall curve (AUPRC) of 0.966 and 0.615 on the trainingand test period, respectively. However, no sufficiently good results were reached with the One-Class Support Vector Machine. The Auto-encoder was made more sophisticated to see how much performance could be increased. It was found that the AUPRC could be increased to 0.987 and 0.801 on the trainingand test period, respectively. Additionally, the model was able to detect and generate one alarm for each incident period that occurred under the test period.The conclusion is that ML can successfully be utilized to detect faults at an earlier stage and potentially circumvent otherwise costly unplanned maintenance. Nevertheless, there is still a lot of room for improvements in the model and the collection of the data.<br>I hopp om att öka identifieringsgraden av störningar i kraftvärmepannor och därigenom minska oplanerat underhåll konstrueras och evalueras tre maskininlärningsmodeller.Algoritmerna; k-Nearest Neighbor, One-Class Support Vector Machine, och Autoencoder har bevisad framgång inom forskning av anomalidetektion, men är relativt outforskade för industriella applikationer som denna på grund av svårigheten att samla in icke-artificiell uppmärkt data inom området.Grundversionerna av k-Nearest Neighbor och Auto-encoder presterade nästan likvärdigt. Dock var Auto-encoder-modellen lite bättre och nådde ett AUPRC-värde av 0.966 respektive 0.615 på träningsoch testperioden. Inget tillräckligt bra resultat nåddes med One-Class Support Vector Machine. Auto-encoder-modellen gjordes mer sofistikerad för att se hur mycket prestandan kunde ökas. Det visade sig att AUPRC-värdet kunde ökas till 0.987 respektive 0.801 under träningsoch testperioden. Dessutom lyckades modellen identifiera och generera ett larm vardera för alla incidenter under testperioden. Slutsatsen är att ML framgångsrikt kan användas för att identifiera störningar iett tidigare skede och därigenom potentiellt kringgå i annat fall dyra oplanerade underhåll. Emellertid finns det fortfarande mycket utrymme för förbättringar av modellen samt inom insamlingen av data.
11
Green, Michael. "Motivation and Learning of Non-Traditional Computing Education Students in a Web-Based Combined Laboratory." Thesis, Nova Southeastern University, 2015. http://pqdtopen.proquest.com/#viewpdf?dispub=3715421.
Full textAbstract:
<p> Hands-on experiential learning activities are an important component of computing education disciplines. Laboratory environments provide learner access to real world equipment for completing experiments. Local campus facilities are commonly used to host laboratory classes. While campus facilities afford hands-on experience with real equipment high maintenance costs, restricted access, and limited flexibility diminish laboratory effectiveness. Web-based simulation and remote laboratory formats have emerged as low cost options, which allow open access and learner control. Simulation lacks fidelity and remote laboratories are considered too complex for novice learners.</p><p> A web-based combined laboratory format incorporates the benefits of each format while mitigating the shortcomings. Relatively few studies have examined the cognitive benefits of web-based laboratory formats in meeting computing education students’ goals. A web-based combined laboratory model that incorporates motivation strategies was developed to address non-traditional computing education students’ preferences for control of pace and access to learning. Internal validation of the laboratory model was conducted using pilot studies and Delphi expert review techniques. A panel of instructors from diverse computing education backgrounds reviewed the laboratory model. Panel recommendations guided enhancement of the model design.</p>
12
Faqeehi, Muhammad Ali. "Exploring the effects of combined strategies on English vocabulary learning among Saudi Arabian university freshmen." Virtual Press, 2003. http://liblink.bsu.edu/uhtbin/catkey/1263924.
Full text
13
Green, Michael Jesse. "Motivation and Learning of Non-Traditional Computing Education Students in a Web-based Combined Laboratory." NSUWorks, 2015. http://nsuworks.nova.edu/gscis_etd/57.
Full textAbstract:
Hands-on experiential learning activities are an important component of computing education disciplines. Laboratory environments provide learner access to real world equipment for completing experiments. Local campus facilities are commonly used to host laboratory classes. While campus facilities afford hands-on experience with real equipment high maintenance costs, restricted access, and limited flexibility diminish laboratory effectiveness. Web-based simulation and remote laboratory formats have emerged as low cost options, which allow open access and learner control. Simulation lacks fidelity and remote laboratories are considered too complex for novice learners.
A web-based combined laboratory format incorporates the benefits of each format while mitigating the shortcomings. Relatively few studies have examined the cognitive benefits of web-based laboratory formats in meeting computing education students’ goals. A web-based combined laboratory model that incorporates motivation strategies was developed to address non-traditional computing education students’ preferences for control of pace and access to learning. Internal validation of the laboratory model was conducted using pilot studies and Delphi expert review techniques. A panel of instructors from diverse computing education backgrounds reviewed the laboratory model. Panel recommendations guided enhancement of the model design.
14
West, Brianna Rose. "Hyperspectral imagery combined with machine learning to differentiate genetically modified (GM) and non-GM canola." Thesis, West, Brianna Rose (2022) Hyperspectral imagery combined with machine learning to differentiate genetically modified (GM) and non-GM canola. Masters by Research thesis, Murdoch University, 2022. https://researchrepository.murdoch.edu.au/id/eprint/64579/.
Full textAbstract:
Canola, also known as rapeseed (Brassica napus L.), is an oilseed that produces a healthy food-grade oil, canola meal by-product, and biofuel. It is the fourth most grown grain in Australia. Genetically modified (GM) canola currently represents approximately twenty percent of national canola production; hence, with clashing public and industry perceptions of genetically modified organisms (GMOs), transparency and traceability must be enabled throughout the supply chain to protect markets and relationships with consumers. GM canola must not cross-contaminate non-GM canola as our largest export market, Europe, has extremely strict protocols on GMOs. GM and non-GM canola cannot be differentiated by the human eye, with polymerase chain reaction (PCR) methods currently the main alternative, which is expensive and time-consuming. This thesis evaluates the potential to differentiate GM from non-GM canola using the novel, rapid, and non-destructive technique of hyperspectral imaging combined with machine learning.
Hyperspectral imagery captures and processes wavelengths beyond simply red, green, and blue. It has a pre-existing multitude of uses including the characterisation and variety identification of other grains. In this study 500 images each of non-GM and GM canola seeds were captured. Seeds were placed on a black background with two lights sources. Images were captured from the 400nm to 1000nm wavelengths, a total of 80 bands, at a 25-millisecond exposure time. These images were run through a convolutional neural network in Keras for analysis. The high dynamic range and raw files were combined into a NumPy file for the hyperspectral image generator. Contrary to expectations, however, the models using the bitmap image files performed similarly to the models receiving the hyperspectral images. Regardless, both produced high validation accuracies around 90%, indicating a detectable phenotypical difference between the two, and further studies could lead to the development of a new approach to GM canola detection.
15
Grancharova, Mila, and Michaela Jangefalk. "Comparative Study of the Combined Performance of Learning Algorithms and Preprocessing Techniques for Text Classification." Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-241435.
Full textAbstract:
With the development in the area of machine learning, society has become more dependent on applications that build on machine learning techniques. Despite this, there are extensive classification tasks which are still performed by humans. This is time costly and often results in errors. One application in machine learning is text classification which has been researched a lot the past twenty years. Text classification tasks can be automated through the machine learning technique supervised learning which can lead to increased performance compared to manual classification. When handling text data, the data often has to be preprocessed in different ways to assure a good classification. Preprocessing techniques have been shown to increase performance of text classification through supervised learning. Different processing techniques affect the performance differently depending on the choice of learning algorithm and characteristics of the data set. This thesis investigates how classification accuracy is affected by different learning algorithms and different preprocessing techniques for a specific customer feedback data set. The researched algorithms are Naïve Bayes, Support Vector Machine and Decision Tree. The research is done by experiments with dependency on algorithm and combinations of preprocessing techniques. The results show that spelling correction and removing stop words increase the accuracy for all classifiers while stemming lowers the accuracy for all classifiers. Furthermore, Decision Tree was most positively affected by preprocessing while Support Vector Machine was most negatively affected. A deeper study on why the preprocessing techniques affected the algorithms in such a way is recommended for future work.<br>I och med utvecklingen inom området maskininlärning har samhället blivit mer beroende av applikationer som bygger på maskininlärningstekniker. Trots detta finns omfattande klassificeringsuppgifter som fortfarande utförs av människor. Detta är tidskrävande och resulterar ofta i olika typer av fel. En uppgift inom maskininlärning är textklassificering som har forskats mycket i de senaste tjugo åren. Textklassificering kan automatiseras genom övervakad maskininlärningsteknik vilket kan leda till effektiviseringar jämfört med manuell klassificering. Ofta måste textdata förbehandlas på olika sätt för att säkerställa en god klassificering. Förbehandlingstekniker har visat sig öka textklassificeringens prestanda genom övervakad inlärning. Olika förbetningstekniker påverkar prestandan olika beroende på valet av inlärningsalgoritm och egenskaper hos datamängden. Denna avhandling undersöker hur klassificeringsnoggrannheten påverkas av olika inlärningsalgoritmer och olika förbehandlingstekniker för en specifik datamängd som utgörs av kunddata. De undersökta algoritmerna är naïve Bayes, supportvektormaskin och beslutsträd. Undersökningen görs genom experiment med beroende av algoritm och kombinationer av förbehandlingstekniker. Resultaten visar att stavningskorrektion och borttagning av stoppord ökar noggrannheten för alla klassificerare medan stämming sänker noggrannheten för alla. Decision Tree var dessutom mest positivt påverkad av de olika förbehandlingsmetoderna medan Support Vector Machine påverkades mest negativt. En djupare studie om varför förbehandlingsresultaten påverkat algoritmerna på ett sådant sätt rekommenderas för framtida arbete.
16
Tumati, Saini. "A Combined Approach to Handle Multi-class Imbalanced Data and to Adapt Concept Drifts using Machine Learning." University of Cincinnati / OhioLINK, 2021. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1623240328088387.
Full text
17
Gupte, Sanket. "A model for strategic bidding in combined transmission and wholesale energy markets." [Tampa, Fla.] : University of South Florida, 2004. http://purl.fcla.edu/fcla/etd/SFE0000391.
Full text
18
Challa, Harshitha. "Multivariate Time-Series Data Requirements in Deep Learning Models." University of Cincinnati / OhioLINK, 2021. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1626356774254081.
Full text
19
Babayan, Bénédicte. "Unraveling the neural circuitry of sequence-based navigation using a combined fos imaging and computational approach." Thesis, Paris 5, 2014. http://www.theses.fr/2014PA05T059/document.
Full textAbstract:
La navigation spatiale est une fonction complexe qui nécessite de combiner des informations sur l’environnement et notre mouvement propre pour construire une représentation du monde et trouver le chemin le plus direct vers notre but. Cette intégration multimodale suggère qu’un large réseau de structures corticales et sous-corticales interagit avec l’hippocampe, structure clé de la navigation. Je me suis concentrée chez la souris sur la navigation de type séquence (ou stratégie égocentrique séquentielle) qui repose sur l’organisation temporelle de mouvements associés à des points de choix spatialement distincts. Après avoir montré que l’apprentissage de cette navigation de type séquence nécessitait l’hippocampe et le striatum dorso-médian, nous avons caractérisé le réseau fonctionnel la sous-tendant en combinant de l’imagerie Fos, de l’analyse de connectivité fonctionnelle et une approche computationnelle. Les réseaux fonctionnels changent au cours de l’apprentissage. Lors de la phase précoce, le réseau impliqué comprend un ensemble de régions cortico-striatales fortement corrélées. L’hippocampe était activé ainsi que des structures impliquées dans le traitement d’informations de mouvement propre (cervelet), dans la manipulation de représentations mentales de l’espace (cortex rétrosplénial, pariétal, entorhinal) et dans la planification de trajectoires dirigées vers un but (boucle cortex préfrontal-ganglions de la base). Le réseau de la phase tardive est caractérisé par l’apparition d’activations coordonnées de l’hippocampe et du cervelet avec le reste du réseau. Parallèlement, nous avons testé si l’intégration de chemin, de l’apprentissage par renforcement basé modèle ou non-basé modèle pouvaient reproduire le comportement des souris. Seul un apprentissage par renforcement non-basé modèle auquel une mémoire rétrospective était ajoutée pouvait reproduire les dynamiques d’apprentissage à l’échelle du groupe ainsi que la variabilité individuelle. Ces résultats suggèrent qu’un modèle d’apprentissage par renforcement suffit à l’apprentissage de la navigation de type séquence et que l’ensemble des structures que cet apprentissage requiert adaptent leurs interactions fonctionnelles au cours de l’apprentissage<br>Spatial navigation is a complex function requiring the combination of external and self-motion cues to build a coherent representation of the external world and drive optimal behaviour directed towards a goal. This multimodal integration suggests that a large network of cortical and subcortical structures interacts with the hippocampus, a key structure in navigation. I have studied navigation in mice through this global approach and have focused on one particular type of navigation, which consists in remembering a sequence of turns, named sequence-based navigation or sequential egocentric strategy. This navigation specifically relies on the temporal organization of movements at spatially distinct choice points. We first showed that sequence-based navigation learning required the hippocampus and the dorsomedial striatum. Our aim was to identify the functional network underlying sequence-based navigation using Fos imaging and computational approaches. The functional networks dynamically changed across early and late learning stages. The early stage network was dominated by a highly inter-connected cortico-striatal cluster. The hippocampus was activated alongside structures known to be involved in self-motion processing (cerebellar cortices), in mental representation of space manipulations (retrosplenial, parietal, entorhinal cortices) and in goal-directed path planning (prefrontal-basal ganglia loop). The late stage was characterized by the emergence of correlated activity between the hippocampus, the cerebellum and the cortico-striatal structures. Conjointly, we explored whether path integration, model-based or model-free reinforcement learning algorithms could explain mice’s learning dynamics. Only the model-free system, as long as a retrospective memory component was added to it, was able to reproduce both the group learning dynamics and the individual variability observed in the mice. These results suggest that a unique model-free reinforcement learning algorithm was sufficient to learn sequence-based navigation and that the multiple structures this learning required adapted their functional interactions across learning
20
Hassan, Mohamed Elhafiz. "Power Plant Operation Optimization : Unit Commitment of Combined Cycle Power Plants Using Machine Learning and MILP." Thesis, mohamed-ahmed@siemens.com, 2019. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-395304.
Full textAbstract:
In modern days electric power systems, the penetration of renewable resources and the introduction of free market principles have led to new challenges facing the power producers and regulators. Renewable production is intermittent which leads to fluctuations in the grid and requires more control for regulators, and the free market principle raises the challenge for power plant producers to operate their plants in the most profitable way given the fluctuating prices. Those problems are addressed in the literature as the Economic Dispatch, and they have been discussed from both regulator and producer view points. Combined Cycle Power plants have the privileges of being dispatchable very fast and with low cost which put them as a primary solution to power disturbance in grid, this fast dispatch-ability also allows them to exploit price changes very efficiently to maximize their profit, and this sheds the light on the importance of prices forecasting as an input for the profit optimization of power plants. In this project, an integrated solution is introduced to optimize the dispatch of combined cycle power plants that are bidding for electricity markets, the solution is composed of two models, the forecasting model and the optimization model. The forecasting model is flexible enough to forecast electricity and fuel prices for different markets and with different forecasting horizons. Machine learning algorithms were used to build and validate the model, and data from different countries were used to test the model. The optimization model incorporates the forecasting model outputs as inputs parameters, and uses other parameters and constraints from the operating conditions of the power plant as well as the market in which the plant is selling. The power plant in this mode is assumed to satisfy different demands, each of these demands have corresponding electricity price and cost of energy not served. The model decides which units to be dispatched at each time stamp to give out the maximum profit given all these constraints, it also decides whether to satisfy all the demands or not producing part of each of them.
21
Bennett, Ronald W. "A comparison of combined and separate text-picture combinations on recall and application of information." Thesis, National Library of Canada = Bibliothèque nationale du Canada, 1997. http://www.collectionscanada.ca/obj/s4/f2/dsk2/ftp04/mq25822.pdf.
Full text
22
Linsdell, Meghan Ashley. "Continuous theta burst stimulation combined with skilled motor practice after stroke : effects on implicit learning and electroneurophysiology." Thesis, University of British Columbia, 2010. http://hdl.handle.net/2429/30475.
Full textAbstract:
After stroke, cortical excitability is decreased in the ipsilesional primary motor cortex and
increased in the contralesional primary motor cortex. This abnormal pattern of excitability
detrimentally affects performance with the hemiparetic arm. Short lasting improvements in
motor performance occur following repetitive transcranial magnetic stimulation (rTMS) over the
contralesional hemisphere after stroke; however, no work has considered the impact of pairing
rTMS with skilled motor practice over multiple days on motor learning, hemiparetic arm
function, or electroneurophysiology in the brain. The aim of this thesis was to determine the
impact of 3 days of continuous theta burst stimulation (cTBS) over contralesional primary motor
cortex paired with skilled motor practice on 1) learning of a novel motor task and hemiparetic
arm motor function and 2) levels of intracortical inhibition, intracortical facilitation, and
transcallosal inhibition following stroke. In a cross-over design, participants with chronic stroke
were randomized to first receive either active or sham cTBS over the contralesional primary
motor cortex. Functional measures, motor task performance, and electroneurophysiology were
assessed at baseline. 3 days of cTBS paired with skilled motor practice were completed;
functional measures, motor learning, and electroneurophysiology were re-evaluated at posttesting.
After a 2-week washout period participants underwent the second half of the study with
the other form of cTBS. Participants showed larger motor learning related change following
active cTBS than sham cTBS. The magnitude of this improvement correlated with enhanced
performance on standardized measures of arm function after stroke. Active cTBS also decreased
levels of facilitation in the contralesional hemisphere and decreased the amount of inhibition
being sent from the contralesional hemisphere to the ipsilesional hemisphere. No adverse effects
were reported. Results of this thesis suggest that cTBS over the contralesional motor cortex
iii
paired with skilled motor practice facilitates both improved hemiparetic arm function and motor
learning beyond that seen with skilled motor practice alone. The results of this thesis contribute
to research relevant to rehabilitation of individuals with stroke and may facilitate the
development of new rehabilitation strategies to improve functional recovery after stroke.
23
Kubacki, Jens [Verfasser]. "Learning and detecting objects in combined range and color images based on local feature frames / Jens Kubacki." Bielefeld : Universitätsbibliothek Bielefeld, 2012. http://d-nb.info/1022199129/34.
Full text
24
Ansved, Linn, and Karin Eklann. "Exploring ways to convey medical information during digital triage : A combined user research and machine learning approach." Thesis, Uppsala universitet, Avdelningen för visuell information och interaktion, 2019. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-386420.
Full textAbstract:
The aim of this project was to investigate what information is critical to convey to nurses when performing digital triage. In addition, the project aimed to investigate how such information could be visualized. This was done through a combined user research and machine learning approach, which enabled for a more nuanced and thorough investigation compared to only making use of one of the two fields. There is sparse research investigating how digital triaging can be improved and made more efficient. Therefore, this study has contributed with new and relevant insights. Three machine learning algorithms were implemented to predict the right level of care for a patient. Out of these three, the random forest classifier proved to have the best performance with an accuracy of 69.46%, also having the shortest execution time. Evaluating the random forest classifier, the most important features were stated to be the duration and progress of the symptoms, allergies to medicine, chronic diseases and the patient's own estimation of his/her health. These factors could all be confirmed by the user research approach, indicating that the results from the approaches were aligned. The results from the user research approach also showed that the patients' own description of their symptoms was of great importance. These findings served as a basis for a number of visualization decisions, aiming to make the triage process as accurate and efficient as possible.
25
Taylor, Carol A. "The Effects of a Combined Mathematics Content/Methods Course Involving Learning Cycles on Preservice Elementary Teachers' Beliefs about the Nature, Teaching and Learning of Mathematics." Diss., Temple University Libraries, 2009. http://cdm16002.contentdm.oclc.org/cdm/ref/collection/p245801coll10/id/22535.
Full textAbstract:
CITE/Mathematics and Science Education<br>Ed.D.<br>The purpose of this study was to determine whether a one-semester combined mathematics content/methods course based on Simon's (1994) Learning Cycle framework for the mathematical preparation of preservice elementary teachers was more effective at aligning preservice teachers' beliefs with the philosophy of the Standards documents (NCTM, 1989, 1991, 1995, 2000) and other reform literature than another content/methods course. While the six-cycled Learning Cycles framework has been used in separate content and methods courses (Simon, 1995), it has not been used in a combined content/methods setting.
The preservice teachers enrolled in two sections of the course Teaching Math to Children: N-6 were involved in the study with one section investigating the topics measurement, geometry and fractions using the Learning Cycles framework while the other section used a different framework to investigate number concepts and place value; addition and subtraction strategies; multiplication and division strategies; geometry and spatial sense; ratio, proportion, percent, similarity and scale; and probability, counting techniques and statistics. Beliefs about the nature, teaching and learning of mathematics were measured both quantitatively and qualitatively. All participants' beliefs were assessed quantitatively at the beginning and the end of the course using a beliefs instrument with three subscales that measured beliefs about the nature, teaching and learning of mathematics. Five preservice teachers from the Learning Cycles course served as case studies and provided qualitative data through interviews and journal prompts.
Analyses of Covariance were calculated for course of enrollment and each of the three beliefs subscales. For beliefs about the nature of mathematics, there was a significant difference between the adjusted post-test scores of the Learning Cycles group and the comparison group favoring the Learning Cycles group. Similar results were found regarding beliefs about the teaching of mathematics. While significant differences were found between the groups with regards to beliefs about the learning of mathematics, the scores on the corresponding subscale changed only slightly. Qualitative analyses confirmed the findings regarding the beliefs about the nature and teaching of mathematics. Differences were also found in the preservice teachers' beliefs about learning mathematics. The case study participants attributed certain characteristics of the course with their changing beliefs. These results indicate that the Learning Cycles course was somewhat successful at aligning preservice teachers' beliefs with those of the reform movement.<br>Temple University--Theses
26
Tou, Li Esther. "Joy of reading : a combined approach to reading acquisition in special classes - children with learning disabilities in Israel." Thesis, Anglia Ruskin University, 2000. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.323521.
Full text
27
AWAN, ZUBAIR ANWAR. "De novo liquid biopsy and radio genomic diagnostic approach with combined deep learning artificial neural networks for NSCLC." Doctoral thesis, Università degli Studi di Palermo, 2022. https://hdl.handle.net/10447/564144.
Full textAbstract:
Each year, the mortality rate and incidence of non-small cell lung cancer (NSCLC) are
dramatically increasing. The introduction of liquid biopsy in the clinical practice of NSCLC
has completely revolutionized the approach to such neoplasm since is generally detected
through complex and invasive procedures and unfortunately at advanced stages. The
importance and innovation of liquid biopsy are linked with the possibility of cancer detection
at every stage, adjuvant treatment, resistance genotyping, systematic initiation of treatment,
minimal residual disease, early detection of relapse, and screening of NSCLC. Circulating
tumor DNA (ctDNA) is now emerging as a non-invasive biomarker that will help to track
tumor burden and allow the monitoring of cancer genome in blood across several malignancies.
Recently, the combination of liquid biopsy and radiomics seems to deliver an efficient way to
study cancer evolution over time providing an important support tool to daily clinical practice.
CT (Computed Tomography) images are of particular importance in this context because they
convey functional and anatomical information, respectively. Machine learning provides a
variety of approaches for dealing with this potentially high-dimensional challenge. In
particular, we used Enet neural network for image assessment. This study represents an
interesting attempt to explore the usefulness of liquid biopsy, radiomics, and deep learning in
the NSCLC clinical routine. We studied a NSCLC patient cohort from the first access to our
department to follow-up. Our results showed a promising correlation between the ctDNA
quantity and radiomic features evaluated by automated computed tomography according to
RECIST criteria with the Enet deep learning method, which allowed us to define more
accurately progression-free survival (PFS) and overall survival (OS) of patients during the
course of cancer history. Therefore, the above mentioned diagnostic tools including the
combination of liquid biopsy, radiomics, and deep learning tools collectively can represent a
very robust and new approach in the monitoring and management of NSCLC.<br>Each year, the mortality rate and incidence of non-small cell lung cancer (NSCLC) are
dramatically increasing. The introduction of liquid biopsy in the clinical practice of NSCLC
has completely revolutionized the approach to such neoplasm since is generally detected
through complex and invasive procedures and unfortunately at advanced stages. The
importance and innovation of liquid biopsy are linked with the possibility of cancer detection
at every stage, adjuvant treatment, resistance genotyping, systematic initiation of treatment,
minimal residual disease, early detection of relapse, and screening of NSCLC. Circulating
tumor DNA (ctDNA) is now emerging as a non-invasive biomarker that will help to track
tumor burden and allow the monitoring of cancer genome in blood across several malignancies.
Recently, the combination of liquid biopsy and radiomics seems to deliver an efficient way to
study cancer evolution over time providing an important support tool to daily clinical practice.
CT (Computed Tomography) images are of particular importance in this context because they
convey functional and anatomical information, respectively. Machine learning provides a
variety of approaches for dealing with this potentially high-dimensional challenge. In
particular, we used Enet neural network for image assessment. This study represents an
interesting attempt to explore the usefulness of liquid biopsy, radiomics, and deep learning in
the NSCLC clinical routine. We studied a NSCLC patient cohort from the first access to our
department to follow-up. Our results showed a promising correlation between the ctDNA
quantity and radiomic features evaluated by automated computed tomography according to
RECIST criteria with the Enet deep learning method, which allowed us to define more
accurately progression-free survival (PFS) and overall survival (OS) of patients during the
course of cancer history. Therefore, the above mentioned diagnostic tools including the
combination of liquid biopsy, radiomics, and deep learning tools collectively can represent a
very robust and new approach in the monitoring and management of NSCLC.
28
Dahab, Sarah. "An approach to measuring software systems using new combined metrics of complex test." Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLL015/document.
Full textAbstract:
La plupart des métriques de qualité logicielle mesurables sont actuellement basées sur des mesures bas niveau, telles que la complexité cyclomatique, le nombre de lignes de commentaires ou le nombre de blocs dupliqués. De même, la qualité de l'ingénierie logicielle est davantage liée à des facteurs techniques ou de gestion, et devrait fournir des indicateurs utiles pour les exigences de qualité. Actuellement, l'évaluation de ces exigences de qualité n'est pas automatisée, elle n'est pas validée empiriquement dans des contextes réels et l'évaluation est définie sans tenir compte des principes de la théorie de la mesure. Par conséquent, il est difficile de comprendre où et comment améliorer le logiciel suivant le résultat obtenu. Dans ce domaine, les principaux défis consistent à définir des métriques adéquates et utiles pour les exigences de qualité, les documents de conception de logiciels et autres artefacts logiciels, y compris les activités de test.Les principales problématiques scientifiques abordées dans cette thèse sont les suivantes: définir des mesures et des outils de support pour mesurer les activités d'ingénierie logicielle modernes en termes d'efficacité et de qualité. La seconde consiste à analyser les résultats de mesure pour identifier quoi et comment s'améliorer automatiquement. Le dernier consiste en l'automatisation du processus de mesure afin de réduire le temps de développement. Une telle solution hautement automatisée et facile à déployer constituera une solution révolutionnaire, car les outils actuels ne le prennent pas en charge, sauf pour une portée très limitée<br>Most of the measurable software quality metrics are currently based on low level metrics, such as cyclomatic complexity, number of comment lines or number of duplicated blocks. Likewise, quality of software engineering is more related to technical or management factoid, and should provide useful metrics for quality requirements. Currently the assessment of these quality requirements is not automated, not empirically validated in real contexts, and the assessment is defined without considering principles of measurement theory. Therefore it is difficult to understand where and how to improve the software following the obtained result. In this domain, the main challenges are to define adequate and useful metrics for quality requirements, software design documents and other software artifacts, including testing activities.The main scientific problematic that are tackled in this proposed thesis are the following : defining metrics and its supporting tools for measuring modern software engineering activities with respect to efficiency and quality. The second consists in analyzing measurement results for identifying what and how to improve automatically. The last one consists in the measurement process automation in order to reduce the development time. Such highly automated and easy to deploy solution will be a breakthrough solution, as current tools do not support it except for very limited scope
29
Patterson, Alexis A. "Effectiveness of combined stress control and behaviour management sessions for parents who have a child with a learning disability." Thesis, University of Edinburgh, 2010. http://hdl.handle.net/1842/5618.
Full textAbstract:
Objectives: There is body of evidence that suggests parents who have a child with a learning disability experience increased stress levels. In addition, research has found a bidirectional relationship between parental stress and child challenging behaviour. Other investigations in this area have found parental stress to be a significant predictor of intervention outcome, and it has been suggested that it should be targeted prior to any child focused intervention. The aim of the current study is to examine the effectiveness of an intervention that targets both parental stress and child challenging behaviour, in families who have a child with a learning disability Design & Method: The study employed a questionnaire based, quantitative, within and between groups methodology. The design included two groups, both parents of children with learning disabilities. Group one, attended three sessions targeting parental stress and three on child challenging behaviour and completed measures on parental stress, mood, coping and child behaviour. Group two, received ‘treatment as normal’ and completed the same measures. Results: Correlations were conducted to look at the relationship between parental stress, coping, mood and child behaviour. The results partially supported a relationship between parental stress, child challenging behaviour and coping. The ANOVA revealed a significant decrease in anxiety scores for the intervention group over time. The results provide some support the effectiveness of the sessions in reducing parental stress but not child challenging behaviour. Conclusions: The findings provide some support for the use of a brief intervention for parents who have a child with a learning disability. The results indicate high levels of stress and child challenging behaviour, which has been linked to increased risk of mental health problems in both child and parents. The current intervention may be appropriative for families with lower levels of stress and behaviour problems, and alternative approaches may be more suitable to families with chronic difficulties.
30
Tam, Janice. "Fostering inter-ethnic relations at an urban school, is knowledge forum combined with face-to-face cooperative learning superior to face-to-face cooperative learning alone?" Thesis, National Library of Canada = Bibliothèque nationale du Canada, 2001. http://www.collectionscanada.ca/obj/s4/f2/dsk3/ftp04/MQ58816.pdf.
Full text
31
Dahab, Sarah. "An approach to measuring software systems using new combined metrics of complex test." Electronic Thesis or Diss., Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLL015.
Full textAbstract:
La plupart des métriques de qualité logicielle mesurables sont actuellement basées sur des mesures bas niveau, telles que la complexité cyclomatique, le nombre de lignes de commentaires ou le nombre de blocs dupliqués. De même, la qualité de l'ingénierie logicielle est davantage liée à des facteurs techniques ou de gestion, et devrait fournir des indicateurs utiles pour les exigences de qualité. Actuellement, l'évaluation de ces exigences de qualité n'est pas automatisée, elle n'est pas validée empiriquement dans des contextes réels et l'évaluation est définie sans tenir compte des principes de la théorie de la mesure. Par conséquent, il est difficile de comprendre où et comment améliorer le logiciel suivant le résultat obtenu. Dans ce domaine, les principaux défis consistent à définir des métriques adéquates et utiles pour les exigences de qualité, les documents de conception de logiciels et autres artefacts logiciels, y compris les activités de test.Les principales problématiques scientifiques abordées dans cette thèse sont les suivantes: définir des mesures et des outils de support pour mesurer les activités d'ingénierie logicielle modernes en termes d'efficacité et de qualité. La seconde consiste à analyser les résultats de mesure pour identifier quoi et comment s'améliorer automatiquement. Le dernier consiste en l'automatisation du processus de mesure afin de réduire le temps de développement. Une telle solution hautement automatisée et facile à déployer constituera une solution révolutionnaire, car les outils actuels ne le prennent pas en charge, sauf pour une portée très limitée<br>Most of the measurable software quality metrics are currently based on low level metrics, such as cyclomatic complexity, number of comment lines or number of duplicated blocks. Likewise, quality of software engineering is more related to technical or management factoid, and should provide useful metrics for quality requirements. Currently the assessment of these quality requirements is not automated, not empirically validated in real contexts, and the assessment is defined without considering principles of measurement theory. Therefore it is difficult to understand where and how to improve the software following the obtained result. In this domain, the main challenges are to define adequate and useful metrics for quality requirements, software design documents and other software artifacts, including testing activities.The main scientific problematic that are tackled in this proposed thesis are the following : defining metrics and its supporting tools for measuring modern software engineering activities with respect to efficiency and quality. The second consists in analyzing measurement results for identifying what and how to improve automatically. The last one consists in the measurement process automation in order to reduce the development time. Such highly automated and easy to deploy solution will be a breakthrough solution, as current tools do not support it except for very limited scope
32
Hanson, Coral Kathleen. "The impact of enrollment in a combined reading and learning strategies course on exceptional students' FCAT Reading scale scores." ScholarWorks, 2009. https://scholarworks.waldenu.edu/dissertations/658.
Full textAbstract:
Previous research has shown that because remediation and support replace required and career-defining courses, exceptional students fall behind, ill equipped to act in society as autonomous adults. No Child Left Behind requires reading proficiency, so students failing standardized tests must take remedial courses. Individualized education plans often require support courses. However, there remains an important gap in the literature regarding the usefulness of reading, standardized testing, and leadership research to solve this problem. A class combining reading and support for students with exceptionalities exists at 1 high school. Therefore, the purpose of this study was to explore the combined program to determine whether a specific program intended to meet federal and state performance-based standards affected test scores. This study used a single-group pretest-posttest design to analyze the 2007 and 2008 Florida Comprehensive Assessment Test Reading scale scores of 25 of the 30 students with exceptionalities enrolled in the combined course to determine whether a significant difference existed between these test scores. The paired-sample t test identified a significant difference between pretest and posttest scores, supporting the hypothesis that combining remediation and support increases progress. This study would be an important contribution to the existing literature by providing a viable solution to this problem by offering more opportunities for exceptional students to enroll in courses available to their mainstream peers. It also would enhance social change initiatives by facilitating the graduation and entry into productive adulthood of students with exceptionalities, allowing them to define career interests and remediate deficiencies simultaneously.
33
Rouhafzay, Ghazal. "3D Object Representation and Recognition Based on Biologically Inspired Combined Use of Visual and Tactile Data." Thesis, Université d'Ottawa / University of Ottawa, 2021. http://hdl.handle.net/10393/42122.
Full textAbstract:
Recent research makes use of biologically inspired computation and artificial intelligence as efficient means to solve real-world problems. Humans show a significant performance in extracting and interpreting visual information. In the cases where visual data is not available, or, for example, if it fails to provide comprehensive information due to occlusions, tactile exploration assists in the interpretation and better understanding of the environment. This cooperation between human senses can serve as an inspiration to embed a higher level of intelligence in computational models.
In the context of this research, in the first step, computational models of visual attention are explored to determine salient regions on the surface of objects. Two different approaches are proposed. The first approach takes advantage of a series of contributing features in guiding human visual attention, namely color, contrast, curvature, edge, entropy, intensity, orientation, and symmetry are efficiently integrated to identify salient features on the surface of 3D objects. This model of visual attention also learns to adaptively weight each feature based on ground-truth data to ensure a better compatibility with human visual exploration capabilities. The second approach uses a deep Convolutional Neural Network (CNN) for feature extraction from images collected from 3D objects and formulates saliency as a fusion map of regions where the CNN looks at, while classifying the object based on their geometrical and semantic characteristics. The main difference between the outcomes of the two algorithms is that the first approach results in saliencies spread over the surface of the objects while the second approach highlights one or two regions with concentrated saliency. Therefore, the first approach is an appropriate simulation of visual exploration of objects, while the second approach successfully simulates the eye fixation locations on objects.
In the second step, the first computational model of visual attention is used to determine scattered salient points on the surface of objects based on which simplified versions of 3D object models preserving the important visual characteristics of objects are constructed. Subsequently, the thesis focuses on the topic of tactile object recognition, leveraging the proposed model of visual attention. Beyond the sensor technologies which are instrumental in ensuring data quality, biological models can also assist in guiding the placement of sensors and support various selective data sampling strategies that allow exploring an object’s surface faster. Therefore, the possibility to guide the acquisition of tactile data based on the identified visually salient features is tested and validated in this research. Different object exploration and data processing approaches were used to identify the most promising solution.
Our experiments confirm the effectiveness of computational models of visual attention as a guide for data selection for both simplifying 3D representation of objects as well as enhancing tactile object recognition. In particular, the current research demonstrates that: (1) the simplified representation of objects by preserving visually salient characteristics shows a better compatibility with human visual capabilities compared to uniformly simplified models, and (2) tactile data acquired based on salient visual features are more informative about the objects’ characteristics and can be employed in tactile object manipulation and recognition scenarios.
In the last section, the thesis addresses the issue of transfer of learning from vision to touch. Inspired from biological studies that attest similarities between the processing of visual and tactile stimuli in human brain, the thesis studies the possibility of transfer of learning from vision to touch using deep learning architectures and proposes a hybrid CNN that handles both visual and tactile object recognition.
34
Weese, Joshua L. "A convolutive model for polyphonic instrument identification and pitch detection using combined classification." Thesis, Kansas State University, 2013. http://hdl.handle.net/2097/15599.
Full textAbstract:
Master of Science<br>Department of Computing and Information Sciences<br>William H. Hsu<br>Pitch detection and instrument identification can be achieved with relatively high accuracy when considering monophonic signals in music; however, accurately classifying polyphonic signals in music remains an unsolved research problem. Pitch and instrument classification is a subset of Music Information Retrieval (MIR) and automatic music transcription, both having numerous research and real-world applications. Several areas of research are covered in this thesis, including the fast Fourier transform, onset detection, convolution, and filtering. Basic music theory and terms are also presented in order to explain the context and structure of data used. The focus of this thesis is on the representation of musical signals in the frequency domain. Polyphonic signals with many different voices and frequencies can be exceptionally complex. This thesis presents a new model for representing the spectral structure of polyphonic signals: Uniform MAx Gaussian Envelope (UMAGE). The new spectral envelope precisely approximates the distribution of frequency parts in the spectrum while still being resilient to oscillating rapidly (noise) and is able to generalize well without losing the representation of the original spectrum. When subjectively compared to other spectral envelope methods, such as the linear predictive coding envelope method and the cepstrum envelope method, UMAGE is able to model high order polyphonic signals without dropping partials (frequencies present in the signal). In other words, UMAGE is able to model a signal independent of the signal’s periodicity. The performance of UMAGE is evaluated both objectively and subjectively. It is shown that UMAGE is robust at modeling the distribution of frequencies in simple and complex polyphonic signals. Combined classification (combiners), a methodology for learning large concepts, is used to simplify the learning process and boost classification results. The output of each learner is then averaged to get the final result. UMAGE is less accurate when identifying pitches; however, it is able to achieve accuracy in identifying instrument groups on order-10 polyphonic signals (ten voices), which is competitive with the current state of the field.
35
Antony, Bhavna Josephine. "A combined machine-learning and graph-based framework for the 3-D automated segmentation of retinal structures in SD-OCT images." Diss., University of Iowa, 2013. https://ir.uiowa.edu/etd/4944.
Full textAbstract:
Spectral-domain optical coherence tomography (SD-OCT) is a non-invasive imaging modality that allows for the quantitative study of retinal structures. SD-OCT has begun to find widespread use in the diagnosis and management of various ocular diseases. While commercial scanners provide limited analysis of a small number of retinal layers, the automated segmentation of retinal layers and other structures within these volumetric images is quite a challenging problem, especially in the presence of disease-induced changes.
The incorporation of a priori information, ranging from qualitative assessments of the data to automatically learned features, can significantly improve the performance of automated methods. Here, a combined machine learning-based approach and graph-theoretic approach is presented for the automated segmentation of retinal structures in SD-OCT images. Machine-learning based approaches are used to learn textural features from a training set, which are then incorporated into the graph- theoretic approach. The impact of the learned features on the final segmentation accuracy of the graph-theoretic approach is carefully evaluated so as to avoid incorporating learned components that do not improve the method. The adaptability of this versatile combination of a machine-learning and graph-theoretic approach is demonstrated through the segmentation of retinal surfaces in images obtained from humans, mice and canines.
In addition to this framework, a novel formulation of the graph-theoretic approach is described whereby surfaces with a disruption can be segmented. By incorporating the boundary of the "hole" into the feasibility definition of the set of surfaces, the final result consists of not only the surfaces but the boundary of the hole as well. Such a formulation can be used to model the neural canal opening (NCO) in SD-OCT images, which appears as a 3-D planar hole disrupting the surfaces in its vicinity. A machine-learning based approach was also used here to learn descriptive features of the NCO.
Thus, the major contributions of this work include 1) a method for the automated correction of axial artifacts in SD-OCT images, 2) a combined machine-learning and graph-theoretic framework for the segmentation of retinal surfaces in SD-OCT images (applied to humans, mice and canines), 3) a novel formulation of the graph-theoretic approach for the segmentation of multiple surfaces and their shared hole (applied to the segmentation of the neural canal opening), and 4) the investigation of textural markers that could precede structural and functional change in degenerative retinal diseases.
36
Sousa, Rafael Nascimento. "The use of slides editor combined with chemistry experimental practice as an educational tool to maximize the teaching and learning process of electrochemical." Universidade Federal do CearÃ, 2015. http://www.teses.ufc.br/tde_busca/arquivo.php?codArquivo=16908.
Full textAbstract:
No mundo pÃs-moderno que à influenciado constantemente pela tecnologia, a educaÃÃo e consequentemente o processo de ensino e aprendizagem invariavelmente sofrerà mudanÃas. Logo, torna-se necessÃrio a concepÃÃo de estratÃgias didÃticas alternativas, procurando-se estabelecer vinculaÃÃes entre o uso pedagÃgico do computador e o laboratÃrio de prÃticas cientÃficas, como ferramentas de apoio pedagÃgico ao desenvolvimento do processo de ensino e aprendizagem. A presente pesquisa focou-se na concepÃÃo de estratÃgias de ensino e aprendizagem atravÃs da realizaÃÃo de sessÃes didÃticas de QuÃmica mediada pelo computador e pela experimentaÃÃo. Tal articulaÃÃo se deu atravÃs da uniÃo pedagÃgica dos laboratÃrios de experimentaÃÃo cientÃfica e de informÃtica com o objetivo de facilitar, aos alunos do 2Âano do ensino mÃdio, a aprendizagem de conhecimentos de FÃsico-QuÃmica, especificamente a pilha de Daniel. Tal pesquisa ancorou-se na proposta construcionista de Valente e no advento das tecnologias digitais que oportunizou a aprendizagem mediada pelo computador, proposto por Pierre Levy. Realizou-se a construÃÃo de um software educativo multimÃdia com o fim de simular a prÃtica experimental da pilha de Daniel, enaltecendo-se a construÃÃo colaborativa de conhecimentos, competÃncias e habilidades. Tais aÃÃes foram realizadas com a participaÃÃo de quatorze alunos do 2 ano do ensino mÃdio de uma escola pÃblica do municÃpio do EusÃbio no CearÃ, concebidas em seis etapas: construÃÃo de um software, aula teÃrica-tradicional, aula prÃtica-virtual mediada pelo computador, avaliaÃÃo do software, prÃtica-experimental e teste avaliativo. A abordagem metodolÃgica da pesquisa foi do tipo quali-quantitativa e se caracterizou como pesquisa descritiva. Os dados obtidos por meio dos questionÃrios, e as observaÃÃes permitiram analisar e concluir haver indÃcios de: desenvolvimento da motivaÃÃo, capacidade criativa e apropriaÃÃo do conteÃdo. O experimento da produÃÃo da pilha de Daniel na etapa prÃtica-experimental foi uma maneira de trabalhar de forma mais clara, dinÃmica, colaborativa a ressignificaÃÃo e construÃÃo de conhecimentos. O uso pedagÃgico do software educacional multimÃdia desenvolvido possibilitou aos alunos uma compreensÃo mais eficiente de conceitos, representaÃÃes e fenÃmenos, alÃm de tornar a aprendizagem do conteÃdo pilha de Daniel mais dinÃmica, motivadora e curiosa.<br>In the postmodern world that is constantly influenced by technology, education and consequently the process of teaching and learning invariably suffer changes. Therefore, it is necessary to design alternative teaching strategies, seeking to establish links between the pedagogical use of computer and laboratory scientific practices as tools for pedagogical support to the development of the teaching and learning process. This research focused on the design of teaching and learning strategies by conducting educational sessions of computer-mediated chemistry and experimentation. Such coordination was made through the educational union of scientific experimentation laboratories and computer in order to facilitate the students of the high school 2nd Year, learning of Physical Chemistry of knowledge, specifically the stack of Daniel. Such research was anchored in constructionist proposal for Valente and the advent of digital technologies that provided an opportunity to learn computer-mediated, proposed by Pierre Levy. There was the construction of a multimedia educational software in order to simulate the experimental practice of Daniel's cell, praising up the collaborative construction of knowledge, skills and abilities. Such actions were carried out with the participation of fourteen students of the 2nd year of high school from a public school of Eusebius the city in CearÃ, designed in six steps: building a software, theoretical-traditional classroom, practice-virtual classroom computer-mediated, software evaluation, experimental and practice-evaluation test. The methodological approach of the research was the qualitative and quantitative type, and was characterized as descriptive research. The data obtained through questionnaires, and observations allowed us to analyze and conclude there is evidence of: development of motivation, creativity and appropriation of content. The experiment of Daniel's cell production in practice and experimental step was a way to work more clearly, dynamic, collaborative and reframing the construction of knowledge. The pedagogical use of multimedia educational software developed allowed students a more efficient understanding of concepts, representations and phenomena, and make learning the Daniel cell content more dynamic, motivating and curious.
37
Lu, Han. "Homeostatic structural plasticity of neuronal connectivity in response to external stimulation : a combined study using computer simulations and in vivo experiments." Electronic Thesis or Diss., Strasbourg, 2020. http://www.theses.fr/2020STRAJ059.
Full textAbstract:
Les neurones pyramidaux sont des unités hautement plastiques des circuits neuronaux. Les boutons axonaux et les épines dendritiques forment ou brisentcontinuellement les synapses. La règle d'Hebb, "les neurones qui s’excitent ensemble, se connectent ensemble", a d'abord été proposée pour expliquer la plasticité synaptique. Cependant, de plus en plus d'études suggèrent que les neurones maintiennent leur fréquence de décharge physiologique par régulation homéostasique. De récentes études montrent que le modèle de plasticité structurelle homéostatique (PSH), combinant à la fois la plasticité structurelle et l’homéostasie de la fréquence de décharge, présente une excellente auto-régulation de la stabilité du réseau tout en atteignant les propriétés Hebbiennes. Ainsi, cette thèse explore la possibilité d'utiliser ce modèle pour expliquer la régulation de la plasticité en réponse à une stimulation externe et précise son échelle de temps biologique<br>Pyramidal neurons are highly plastic units of neural circuits. Under basal conditions, axonal boutons and dendritic spines continually form or break synapses. External perturbations further accelerate such changes. Hebb's rule,``fire together, wire together'', was first proposed to explain the observed synaptic plasticity. However, accumulating evidence suggests neurons homeostatically control their neural activity. The conventional solution in computer simulations is combining Hebb's rule with other homeostatic mechanisms to stabilize network dynamics. A recent model of homeostatic structural plasticity (HSP), which includes structural changes and negative feedback control of neural activity, preserves network stability and meanwhile presents similar properties as Hebb's rule. This thesis explores the feasibility of using this model to explain the plasticity induced by external stimulations and pinpoints its time scale
38
MASTRO, PIETRO. "Innovative Techniques for the Retrieval of Earth’s Surface and Atmosphere Geophysical Parameters: Spaceborne Infrared/Microwave Combined Analyses." Doctoral thesis, Università degli studi della Basilicata, 2023. https://hdl.handle.net/11563/162986.
Full textAbstract:
With the advent of the first satellites for Earth Observation: Landsat-1 in July 1972 and ERS-1 in May 1991, the discipline of environmental remote sensing has become, over time, increasingly fundamental for the study of phenomena characterizing the planet Earth. The goal of environmental remote sensing is to perform detailed analyses and to monitor the temporal evolution of different physical phenomena, exploiting the mechanisms of interaction between the objects that are present in an observed scene and the electromagnetic radiation detected by sensors, placed at a distance from the scene, operating at different frequencies. The analyzed physical phenomena are those related to climate change, weather forecasts, global ocean circulation, greenhouse gas profiling, earthquakes, volcanic eruptions, soil subsidence, and the effects of rapid urbanization processes. Generally, remote sensing sensors are of two primary types: active and passive. Active sensors use their own source of electromagnetic radiation to illuminate and analyze an area of interest. An active sensor emits radiation in the direction of the area to be investigated and then detects and measures the radiation that is backscattered from the objects contained in that area. Passive sensors, on the other hand, detect natural electromagnetic radiation (e.g., from the Sun in the visible band and the Earth in the infrared and microwave bands) emitted or reflected by the object contained in the observed scene. The scientific community has dedicated many resources to developing techniques to estimate, study and analyze Earth’s geophysical parameters. These techniques differ for active and passive sensors because they depend strictly on the type of the measured physical quantity. In my P.h.D. work, inversion techniques for estimating Earth’s surface and atmosphere geophysical parameters will be addressed, emphasizing methods based on machine learning (ML). In particular, the study of cloud microphysics and the characterization of Earth’s surface changes phenomenon are the critical points of this work.
39
Guin, Agneev. "Terrain Classification to find Drivable Surfaces using Deep Neural Networks : Semantic segmentation for unstructured roads combined with the use of Gabor filters to determine drivable regions trained on a small dataset." Thesis, KTH, Robotik, perception och lärande, RPL, 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-222021.
Full textAbstract:
Autonomous vehicles face various challenges under difficult terrain conditions such as marginally rural or back-country roads, due to the lack of lane information, road signs or traffic signals. In this thesis, we investigate a novel approach of using Deep Neural Networks (DNNs) to classify off-road surfaces into the types of terrains with the aim of supporting autonomous navigation in unstructured environments. For example, off-road surfaces can be classified as asphalt, gravel, grass, mud, snow, etc. Images from the camera mounted on a mining truck were used to perform semantic segmentation and to classify road surface types. Camera images were segmented manually for training into sets of 16 and 9 classes, for all relevant classes and the drivable classes respectively. A small but diverse dataset of 100 images was augmented and compiled along with nearby frames from the video clips to expand this dataset. Neural networks were used to test the performance for the classification under these off-road conditions. Pre-trained AlexNet was compared to the networks without pre-training. Gabor filters, known to distinguish textured surfaces, was further used to improve the results of the neural network. The experiments show that pre-trained networks perform well with small datasets and many classes. A combination of Gabor filters with pre-trained networks can establish a dependable navigation path under difficult terrain conditions. While the results seem positive for images similar to the training image scenes, the networks fail to perform well in other situations. Though the tests imply that larger datasets are required for dependable results, this is a step closer to making the autonomous vehicles drivable under off-road conditions.<br>Autonoma fordon står inför olika utmaningar under svåra terrängförhållanden som landsbygds- eller skogsvägar på grund av bristen av körfältinformation, vägskyltar och trafikljus. I denna avhandling undersöker vi ett nytt tillvägagångssätt att använda Djupa Neurala Nätverk (DNN) för att klassificera terrängytor utifrån deras körbarhet i syfte att stödja autonom navigering i ostrukturerade miljöer.Till exempel kan terrängytor klassificeras som asfalt, grus, gräs, lera, snö etc. Bilder från kameran monterad på en gruvbil användes för att utföra semantisk segmentering och klassificera vägytor. Bilderna delades manuellt upp i träningsset på 16 samt 9 klasser för alla relevanta klasser respektive körbara klasser. Ett litet men mångsidigt dataset med 100 bilder förstärktes med närliggande bilder från videoklippen för att expandera detta dataset. Neurala nätverk användes för att testa prestandan hos klassificeringen under dessa terrängförhållanden. Det förtränade nätverket AlexNet jämfördes med nätverken utan träning. Gaborfilter, kända för att särskilja texturerade ytor, användes vidare för att förbättra resultaten av det neurala nätverket. Experimenten visar att förtränade nätverk presterar bra med små dataset och många klasser. En kombination av Gaborfilter med förtränade nätverk kan skapa en pålitlig navigationsväg under svåra terrängförhållanden. Även om resultaten verkar positiva för bilder som liknar träningsbildscenen presterar nätverken inte bra i andra situationer. Även om testen tyder på att stora dataset krävs för tillförlitliga resultat, är detta ett steg närmare att göra de autonoma bilarna körbara i svåra terrängförhållanden.
40
Fankhauser, Craig Charles. "A multi-sensory, study strategies and cooperative group instructional approach combined to teach American history to mildly disabled and non disabled adolescents." CSUSB ScholarWorks, 1991. https://scholarworks.lib.csusb.edu/etd-project/577.
Full text
41
Tsai, Po-Yen, and 蔡柏彥. "Teaching Performance Evaluation for Enhancing Multimedia Learning Combined with Learning Path and Cued Retrospective Report." Thesis, 2013. http://ndltd.ncl.edu.tw/handle/25121566029948286483.
Full textAbstract:
碩士<br>國立東華大學<br>資訊工程學系<br>101<br>Many researches indicated that the cued retrospective reporting can increase students’ learning performance, and it is more suitable for the novice. This work aims at establishing a multimedia teaching assistant system combined with cued retrospective reporting provided with learning path to prompt students towards self-learning. With this automatic learning assistant method, we can not only alleviate the teaching load of teachers, but also increase the learning performance of students.
Recently, e-learning becomes popular, but most is lack of the mechanism of automation. With an automation system, teachers’ teaching load can be lowered greatly. However, how to distinguish the right concept for a classifier is an important task. The proposed work intended to utilize the technique of text mining to make classification for students’ posted concepts with the training via one-class Support Vector Machines (SVMs) classifier, and to modify the term weight of keywords. Therefore , the right and wrong concepts can be determined with a high accuracy.
Besides, an expert guidance mechanism can guide the students that are not majoring in some specific topics to increase their learning performance. The experimental results revealed that the proposed Multimedia Teaching and Expert Guidance Mechanism can effectively improve the blind spots on learning process of students.
42
Nag, Abhikesh. "Combined generative-discriminative learning for object recognition using local image descriptors." 2008. http://etda.libraries.psu.edu/theses/approved/PSUonlyIndex/ETD-2506/index.html.
Full text
43
Chen, Shu-ling, and 陳淑玲. "Action Research on Elementary School Senior Students’ Learning Attitude Towards Music Learning Combined With Creative Drama." Thesis, 2004. http://ndltd.ncl.edu.tw/handle/83336792332638055461.
Full textAbstract:
碩士<br>國立臺南大學<br>國民教育研究所<br>92<br>The purpose of this thesis is to investigate the influence of creative drama, when applied to music teaching and learning, upon elementary school senior students’ learning attitude. The author applied action research to investigate the teaching gist and techniques of creative drama, the practical methods when applied to music teaching and learning, and the ways students learned. Subjects were this author’s students, one class from Grade 5, the other from Grade 6. This investigation lasted for 24 classes. A simple research in quantity was added to help the author to understand students’ learning attitude after creative drama education.
The major results are as follow,
Ⅰ.The teaching theory of creative drama:
Creative drama is a classwide activity which involves a guider/teacher to help the participants/students express their imagination in dramatic ways. It is a collaborative impromptu production which puts its emphasis on the process and the pressure-free environment instead of the result.
Ⅱ.When teaching creative drama, a teacher must
1. do teaching planning well, make good use of teaching stragegies, familiarize himeself/herself with guiding process and consulting techniques, make clear explainations and suggestions, fully use of teaching resources, and be reflective with backwash research.
2. have professionalism and passion, be a good listener towards students with encouragments and proper counselling, and accept students'' opinions.
Ⅲ.Practical methods:
1. Use musical games as warm-up activities.
2. Combine body language, mimes, role-playing, and impromptu performance to teach music.
3. The integrity and step-by-step quality of music teaching and learning should not be sacrificed.
4. Teach in a slow but steady way according to the students’ experiences and abilities.
5. Arrange the class as a continuous two-period session.
Ⅳ.Problems encountered during research:
Adjustments in academic knowledge in learning music, time limit, teaching planning, and proper classrooms are problems needed to be solved.
Ⅴ.The influence upon elementary school senior students’ learning attitude:
1. Most students enjoyed themselves more which enhanced their motives to learn.
2. Students were more willing to share their opinions and productions;
3. Classroom interactions were increasing and students experienced more joy in collaborative production;
4. When performing, most students were willing to express themselves.
According the results of this research, the author makes suggestions for music teachers and executive offices in elementary schools.
44
江子揚, Chiang Tzu-Yang, and 江子揚. "Course Design for Elementary School about Woodworking Learning Combined with Environmental Education." Thesis, 2017. http://ndltd.ncl.edu.tw/handle/rp32g4.
Full textAbstract:
碩士<br>國立屏東科技大學<br>木材科學與設計系所<br>105<br>This study is about wood education, which is based on participating in community building, and being taken seriously about environmental education. Therefore, we expect to establish students' environmental awareness and attitude by wood education, and produce culture link with their hometown. For this reason, we conducted teaching experiments in two sixth grade class in Sing-Hua Elementary School. This study has applied the observation method to describe the progress of the courses, and the questionnaire analysis to compare the differences between original and improved courses. It is the way to test the effectiveness of the improved course as a follow-up reference. The results of observation of original course showed: 1. Students prefer to use wire saw and plane cutter, but don’t like to use chisel and graver. 2. It will reduce the willingness to learn if the course was too tight or the frequency of repeatability work was too high. 3. Knowledge content needs to be closer to life, and presented in a simple context. Furthermore, the results of observations by regression analysis show that the correlation coefficient between the evaluation indicators and students learning situation are 0.60, 0.65, and -0.15. It means the factors which affect the students' learning situation are the practicability and difficulty of production. Finally, the results of questionnaire showed: 1. Average values of two courses are 57.53 and 76.63, and P-value of T-test between the two courses is 6.6E-11, which is significant difference. 2. P-value of T-test between the gender and two courses are 0.39 and 0.30, which are no significant difference. 3. The correlation coefficient of satisfaction between knowledge and handmade content in two courses are 0.33 and 0.72. It shows students think that there is a great influence between knowledge and handmade content in improved course.
【Keywords】Elementary School, Woodworking Learning, Environmental Education, Course Design
45
Chiu, Li-yu, and 邱俐瑜. "Inquiry learning combined with a situated educational computer game for improving students' learning performance in financial courses." Thesis, 2014. http://ndltd.ncl.edu.tw/handle/45587281257010052137.
Full textAbstract:
碩士<br>國立臺灣科技大學<br>數位學習與教育研究所<br>102<br>Inquiry learning could enhance students' active learning attitude and higher-order thinking. But for situational content, due to the lack of situation, it's difficult for students to take out to use and solve problems in real life. With a quasi-experiment, the study attempts to applied the system to a financial course. The subjects were assigned to experimental group and control group. Experimental group used inquiry learning combined with a situated educational computer game, and control group used inquiry learning combined with online search engines. In addition, due to the learning style play an important role in learning process, the subjects also been divided into "Active" group and " Reflective" group. The experimental results show that experimental group performed better than control group. The "Active" students in experimental group performed better than others at learning achievement. At learning motivation, learning satisfaction, flow experience and technology acceptance, experimental group is higher than control group. At mental load, experimental group is lower than control group. At self-efficacy and mental efforts, there is no difference between experimental group and control group. Therefore, provided students situated enviromemt to explore can effectively improve their learning performance. In addition, students' learning style should be taken into account, which can better enhance their learning performance.
46
WANG, YE-JHAN, and 王葉展. "A Study of Image Recognition Using Combined Convolutional Neural Networks and Transfer Learning." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/kxb7sv.
Full textAbstract:
碩士<br>國立高雄科技大學<br>資訊管理系<br>107<br>With the growth of GPU performance, more and more people are investing in deep learning. At the same time, due to ImageNet's image recognition competition, many deep network models have been developed. This study explores LeNet-5, AlexNet, VGG16, ResNet50, and Inception. V4, DenseNet121 and other common models and combined models for the assessment of the ability to identify. Three datasets of leaf disease (PlantVillage Disease Leaf), fish (Fish Recognition Ground-Truth data) and personal collection of food materials (Taiwanese foods and drinks) are applied in our study.
In addition to the training using the original models, the combined model of migration learning and pairwise combination and the multiple output of the model are also used. In the experimental results, DenseNet121 performed better than the other models in the test set of the three data sets. In the experiments using transfer learning, the fish and food data sets were not better than the retraining model. Although the training time using transfer learning is only half of the original models but the recognition rate is not ideal. In the experiments of combined models, the combined training of LeNet-5 and VGG16 has higher recognition rate performance than the single model training. The use of leaf disease transfer learning in multiple output experiments is superior to retraining. In the data augmented experiments, the recognition rate of the three data sets can be improved.
In the experiment of this study, the middle layer was extracted from the trained model. The middle layer was visualized as a heat map and was used to observe the state to identify whether the model extracts the corresponding feature when identifying the object, or focuses on the image. The heat map can help us to determine if the model can correctly identify the target.
47
Esposito, Andrea. "Combined polymer physics and machine learning approach to investigate the chromosome 3D structure." Tesi di dottorato, 2020. http://www.fedoa.unina.it/12999/1/Esposito_Andrea_32.pdf.
Full textAbstract:
The spatial organization of the chromatin in the nucleus is known to play an important role in transcriptional regulation of genes in many organisms. However, the comprehension of genome architecture and of the molecular mechanisms shaping its structure, represents a challenging problem which remains not fully understood. During the last two decades, the development of new technologies has allowed to investigate the three-dimensional spatial folding of chromosomes in a quantitative way. Thanks to these technologies, we now know that chromosomes are characterized by a complex, non-random, 3D organization occurring at different genomic length scales, through local and long-range interactions. The molecular factors underlying their formation are still to be investigated. In this sense, polymer physics is turning out to be a great tool to understand the molecular mechanisms of the 3D chromatin spatial organization from first principles. The studies discussed in the present work have been devised in this general framework. They consist of a detailed description of results and conclusions from the projects that we have conducted in the Physics Department of University of Naples Federico II, under the supervision of Professor Mario Nicodemi, in the group of Complex Systems. Many results have been published or are currently in progress in collaboration with the Epigenetic Regulation and Chromatin Architecture group directed by Prof. Ana Pombo, at Max Delbruck Centre For Molecular Medicine (Berlin) and the Development and Disease Group directed by Professor Stefan Mundlos, at Max Planck Institute for Molecular Genetics (Berlin).
The thesis is organized in four chapters. In Chapter 1, we introduce some basic concepts necessary to the comprehension of this research activity and summarize recent results related to the chromatin spatial organization, as the main experimental techniques, the interpretation of the chromosome interaction data and the relationship between spatial organization and cell functionality. Then, we briefly review the polymer models currently proposed to describe the chromosomes three-dimensional organization in the cell nucleus. In Chapter 2, we outline the ‘Strings and Binders Switch (SBS)’ model, developed in our research group, and we make use of it to quantitatively explain the information contained in the Hi-C interaction data via Molecular Dynamics simulations. We show that the thermodynamic phases envisaged by our mode can be used to explain the long-range contact profile of chromosomes; then we try to schematically model the hierarchical structure of chromatin, and finally we present a theoretical study of the multiple co-localization contact landscape. In Chapter 3, we introduce more sophisticated variant of the SBS polymer model by which we can reconstruct the 3D structure of real genomic region with high accuracy. Next, we employ this model to study the folding mechanisms and the enhancer-promoter communication at some important chromosome loci where the failure of these mechanisms can lead to severe diseases. Finally, in Chapter 4, we extend our modeling genome-wide, i.e. to the entire set of chromosomes of the mouse genome. The increase in statistics obtained with the genome-wide study, allows us to compare our polymer models with epigenetics factors, known to play an important role in gene regulation. In this way, we can clarify the molecular nature of the binding factors inferred by our model.
48
CHU, CHIEN-HAO, and 朱健豪. "A Study of PPT Presentation Teaching Combined with Cooperative Learning on the Learning Effectiveness for the Ninth Grade Students." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/snfw6c.
Full textAbstract:
碩士<br>國立高雄科技大學<br>人力資源發展系<br>107<br>The main purpose of this study is to compare the impact of the "PPT presentation teaching combined with cooperative learning" and "PPT presentation teaching" on the achievement of mathematics learning and mathematics learning attitudes for the ninth grade students. This study used a quasi-experimental study of action research. The study was conducted in two classes in the ninth grade of a country in a remote area of Kaohsiung City, with a total of 47 students. With the first summary assessment as the watershed, both classes used the PPT presentation teaching before the first summary assessment. After the first summary assessment, the experimental group implemented the “PPT presentation teaching combined with cooperative learning”, while the control group class still implements the "PPT presentation teaching" to compare students learning achievements, learning attitudes and feelings about group cooperative learning. After the experiment, the results of the test are determined as independent sample t-test and dependent sample t-test. The results of the research are described as follows: 1. After the experimental teaching period, the achievement of the low-score students in the experimental group significantly surpasses that in the control group. On the other hand, the achievement of the middle-score or high-score students in the experimental group is on a par with that in the control group. 2. After the experimental teaching, the learning attitude of the experimental group students is significantly better than that of the control group students. 3. All the students in the experimental group make a significant progress in the attitude of mathematics learning. 4. PPT presentation teaching combined with cooperative learning can definitely enhance the achievement in mathematics learning for the ninth grade students. Based on the above results of the study, this study provides relevant recommendations as a reference for future teaching and related research to the instructor, and researchers.
49
CHANG, HUI-YUN, and 張慧韻. "Effect of Mathematical Game Combined with Polya Strategy on Elementary School Students' Learning Performance." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/e337q3.
Full textAbstract:
碩士<br>國立臺北教育大學<br>數學暨資訊教育學系(含數學教育碩士班)<br>106<br>The two-step word problem is the focus of the lower-level student's learning, and is also the basis for learning multi-step text problems later. However, to solve the two-step word problem requires more integration of the problem, which often leads to students’ learning load. Scholars believe that mathematics learning in primary schools should properly learn “how to solve problems.” With the advancement of information technology, many studies have pointed out that digital game-based learning has great benefits for learning effectiveness, but it should incorporate appropriate teaching strategies. Polya strategy consists of four stages. It is very suitable to assist students in learning mathematics. It is very helpful for schoolchildren to learn a complete course of thinking and problem solving. This study adopts a quasi-experimental design with a total of 42 students in two classes in the second year of elementary school. Classes were divided into experimental group and control group. Both groups of students used mathematics game-based learning. The theme of two-step addition and subtraction in the second grade of the elementary school was the theme. A combination of Polya strategies and role-playing situational game learning was proposed. The experimental group studied through the digital game-based learning system which combined with Polya strategy; the control group learned through the general digital game-based learning system. This study found that the mathematical game-based learning model that combined with Polya strategy can help improve students' learning effectiveness and reduce their math anxiety. Using game-based learning combined with Polya strategy does not reduce students’ Learning motivation and does not increase students’ cognitive load. Students using different learning modes hold a positive attitude towards game-based learning.
50
Wei-Hsiang-Wang and 王韋翔. "Deep learning combined with genetic algorithm for the life cycle prediction of game Apps." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/4v7evy.
Full textAbstract:
碩士<br>元智大學<br>資訊管理學系<br>106<br>Hardware development has led to the growth of deep learning. In recent years, deep learning is popular in many field, especially in speech recognition and image recognition. This study retrieves comments from the web and convert it into meaningful information by using text mining techniques. Then use deep learning to predict the vertices and valleys of the game life cycle of Google Play, and compare the results with traditional machine learning. Finally, combine convolutional neural networks and feature extraction of genetic algorithm to reduce the amount of computation for deep learning and enhance the effects. The results show that the prediction result of convolutional neural network is 97.5%, and the effect of feature extraction by convolutional neural network combined with genetic algorithm reaches 98.44%, and its growth rate is 0.94%.