To see the other types of publications on this topic, follow the link: Recalling-based recurrent neural network.
Dissertations / Theses on the topic 'Recalling-based recurrent neural network'
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 'Recalling-based recurrent neural network.'
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
He, Jian. "Adaptive power system stabilizer based on recurrent neural network." Thesis, National Library of Canada = Bibliothèque nationale du Canada, 1999. http://www.collectionscanada.ca/obj/s4/f2/dsk1/tape8/PQDD_0008/NQ38471.pdf.
Full text
2
Moradi, Mahdi. "TIME SERIES FORECASTING USING DUAL-STAGE ATTENTION-BASED RECURRENT NEURAL NETWORK." OpenSIUC, 2020. https://opensiuc.lib.siu.edu/theses/2701.
Full textAbstract:
AN ABSTRACT OF THE RESEARCH PAPER OFMahdi Moradi, for the Master of Science degree in Computer Science, presented on April 1, 2020, at Southern Illinois University Carbondale.TITLE: TIME SERIES FORECASTING USING DUAL-STAGE ATTENTION-BASED RECURRENT NEURAL NETWORKMAJOR PROFESSOR: Dr. Banafsheh Rekabdar
3
Wang, Yuchen. "Detection of Opioid Addicts via Attention-based bidirectional Recurrent Neural Network." Case Western Reserve University School of Graduate Studies / OhioLINK, 2020. http://rave.ohiolink.edu/etdc/view?acc_num=case1592255095863388.
Full text
4
Wang, Xutao. "Chinese Text Classification Based On Deep Learning." Thesis, Mittuniversitetet, Avdelningen för informationssystem och -teknologi, 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:miun:diva-35322.
Full textAbstract:
Text classification has always been a concern in area of natural language processing, especially nowadays the data are getting massive due to the development of internet. Recurrent neural network (RNN) is one of the most popular method for natural language processing due to its recurrent architecture which give it ability to process serialized information. In the meanwhile, Convolutional neural network (CNN) has shown its ability to extract features from visual imagery. This paper combine the advantages of RNN and CNN and proposed a model called BLSTM-C for Chinese text classification. BLSTM-C begins with a Bidirectional long short-term memory (BLSTM) layer which is an special kind of RNN to get a sequence output based on the past context and the future context. Then it feed this sequence to CNN layer which is utilized to extract features from the previous sequence. We evaluate BLSTM-C model on several tasks such as sentiment classification and category classification and the result shows our model’s remarkable performance on these text tasks.
5
Taylor, Adrian. "Anomaly-Based Detection of Malicious Activity in In-Vehicle Networks." Thesis, Université d'Ottawa / University of Ottawa, 2017. http://hdl.handle.net/10393/36120.
Full textAbstract:
Modern automobiles have been proven vulnerable to hacking by security researchers. By exploiting vulnerabilities in the car's external interfaces, attackers can access a car's controller area network (CAN) bus and cause malicious effects. We seek to detect these attacks on the bus as a last line of defence against automotive cyber attacks. The CAN bus standard defines a low-level message structure, upon which manufacturers layer their own proprietary command protocols; attacks must similarly be tailored for their target. This variability makes intrusion detection methods difficult to apply to the automotive CAN bus. Nevertheless, the bus traffic is generated by machines; thus we hypothesize that it can be characterized with machine learning, and that attacks produce anomalous traffic. Our goals are to show that anomaly detection trained without understanding of the message contents can detect attacks, and to create a framework for understanding how the characteristics of a novel attack can be used to predict its detectability.
We developed a model that describes attacks based on their effect on bus traffic, informed by a review of published material on car hacking in combination with analysis of CAN traffic from a 2012 Subaru Impreza. The model specifies three high-level categories of effects: attacks that insert foreign packets, attacks that affect packet timing, and attacks that only modify data within packets. Foreign packet attacks are trivially detectable. For timing-based anomalies, we developed features suitable for one-class classification methods. For packet stream data word anomalies, we adapted recurrent neural networks and multivariate Markov model methods to sequence anomaly detection and compared their performance.
We conducted experiments to evaluate our detection methods with special attention to the trade-off between precision and recall, given that a practical system requires a very low false alarm rate. The methods were evaluated by synthesizing anomalies within each attack category, parameterized to adjust their covertness. We generalize from the results to enable prediction of detection rates for new attacks using these methods.
6
Zheng, Yilin. "Text-Based Speech Video Synthesis from a Single Face Image." The Ohio State University, 2019. http://rave.ohiolink.edu/etdc/view?acc_num=osu1572168353691788.
Full text
7
Max, Lindblad. "The impact of parsing methods on recurrent neural networks applied to event-based vehicular signal data." Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-223966.
Full textAbstract:
This thesis examines two different approaches to parsing event-based vehicular signal data to produce input to a neural network prediction model: event parsing, where the data is kept unevenly spaced over the temporal domain, and slice parsing, where the data is made to be evenly spaced over the temporal domain instead. The dataset used as a basis for these experiments consists of a number of vehicular signal logs taken at Scania AB. Comparisons between the parsing methods have been made by first training long short-term memory (LSTM) recurrent neural networks (RNN) on each of the parsed datasets and then measuring the output error and resource costs of each such model after having validated them on a number of shared validation sets. The results from these tests clearly show that slice parsing compares favourably to event parsing.<br>Denna avhandling jämför två olika tillvägagångssätt vad gäller parsningen av händelsebaserad signaldata från fordon för att producera indata till en förutsägelsemodell i form av ett neuronnät, nämligen händelseparsning, där datan förblir ojämnt fördelad över tidsdomänen, och skivparsning, där datan är omgjord till att istället vara jämnt fördelad över tidsdomänen. Det dataset som används för dessa experiment är ett antal signalloggar från fordon som kommer från Scania. Jämförelser mellan parsningsmetoderna gjordes genom att först träna ett lång korttidsminne (LSTM) återkommande neuronnät (RNN) på vardera av de skapade dataseten för att sedan mäta utmatningsfelet och resurskostnader för varje modell efter att de validerats på en delad uppsättning av valideringsdata. Resultaten från dessa tester visar tydligt på att skivparsning står sig väl mot händelseparsning.
8
Liu, Chang. "Data Analysis of Minimally-Structured Heterogeneous Logs : An experimental study of log template extraction and anomaly detection based on Recurrent Neural Network and Naive Bayes." Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-191334.
Full textAbstract:
Nowadays, the ideas of continuous integration and continuous delivery are under heavy usage in order to achieve rapid software development speed and quick product delivery to the customers with good quality. During the process ofmodern software development, the testing stage has always been with great significance so that the delivered software is meeting all the requirements and with high quality, maintainability, sustainability, scalability, etc. The key assignment of software testing is to find bugs from every test and solve them. The developers and test engineers at Ericsson, who are working on a large scale software architecture, are mainly relying on the logs generated during the testing, which contains important information regarding the system behavior and software status, to debug the software. However, the volume of the data is too big and the variety is too complex and unpredictable, therefore, it is very time consuming and with great efforts for them to manually locate and resolve the bugs from such vast amount of log data. The objective of this thesis project is to explore a way to conduct log analysis efficiently and effectively by applying relevant machine learning algorithms in order to help people quickly detect the test failure and its possible causalities. In this project, a method of preprocessing and clusering original logs is designed and implemented in order to obtain useful data which can be fed to machine learning algorithms. The comparable log analysis, based on two machine learning algorithms - Recurrent Neural Network and Naive Bayes, is conducted for detecting the place of system failures and anomalies. Finally, relevant experimental results are provided and analyzed.
9
Keisala, Simon. "Using a Character-Based Language Model for Caption Generation." Thesis, Linköpings universitet, Interaktiva och kognitiva system, 2019. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-163001.
Full textAbstract:
Using AI to automatically describe images is a challenging task. The aim of this study has been to compare the use of character-based language models with one of the current state-of-the-art token-based language models, im2txt, to generate image captions, with focus on morphological correctness. Previous work has shown that character-based language models are able to outperform token-based language models in morphologically rich languages. Other studies show that simple multi-layered LSTM-blocks are able to learn to replicate the syntax of its training data. To study the usability of character-based language models an alternative model based on TensorFlow im2txt has been created. The model changes the token-generation architecture into handling character-sized tokens instead of word-sized tokens. The results suggest that a character-based language model could outperform the current token-based language models, although due to time and computing power constraints this study fails to draw a clear conclusion. A problem with one of the methods, subsampling, is discussed. When using the original method on character-sized tokens this method removes characters (including special characters) instead of full words. To solve this issue, a two-phase approach is suggested, where training data first is separated into word-sized tokens where subsampling is performed. The remaining tokens are then separated into character-sized tokens. Future work where the modified subsampling and fine-tuning of the hyperparameters are performed is suggested to gain a clearer conclusion of the performance of character-based language models.
10
He, Fan. "Real-time Process Modelling Based on Big Data Stream Learning." Thesis, Mälardalens högskola, Akademin för innovation, design och teknik, 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:mdh:diva-35823.
Full textAbstract:
Most control systems now are assumed to be unchangeable, but this is an ideal situation. In real applications, they are often accompanied with many changes. Some of changes are from environment changes, and some are system requirements. So, the goal of thesis is to model a dynamic adaptive real-time control system process with big data stream. In this way, control system model can adjust itself using example measurements acquired during the operation and give suggestion to next arrival input, which also indicates the accuracy of states under control highly depends on quality of the process model. In this thesis, we choose recurrent neural network to model process because it is a kind of cheap and fast artificial intelligence. In most of existent artificial intelligence, a database is necessity and the bigger the database is, the more accurate result can be. For example, in case-based reasoning, testcase should be compared with all of cases in database, then take the closer one’s result as reference. However, in neural network, it does not need any big database to support and search, and only needs simple calculation instead, because information is all stored in each connection. All small units called neuron are linear combination, but a neural network made up of neurons can perform some complex and non-linear functionalities. For training part, Backpropagation and Kalman filter are used together. Backpropagation is a widely-used and stable optimization algorithm. Kalman filter is new to gradient-based optimization, but it has been proved to converge faster than other traditional first-order-gradient-based algorithms. Several experiments were prepared to compare new and existent algorithms under various circumstances. The first set of experiments are static systems and they are only used to investigate convergence rate and accuracy of different algorithms. The second set of experiments are time-varying systems and the purpose is to take one more attribute, adaptivity, into consideration.
11
Almejalli, Khaled A. "Intelligent Real-Time Decision Support Systems for Road Traffic Management. Multi-agent based Fuzzy Neural Networks with a GA learning approach in managing control actions of road traffic centres." Thesis, University of Bradford, 2010. http://hdl.handle.net/10454/4264.
Full textAbstract:
The selection of the most appropriate traffic control actions to solve non-recurrent traffic congestion is a complex task which requires significant expert knowledge and experience. In this thesis we develop and investigate the application of an intelligent traffic control decision support system for road traffic management to assist the human operator to identify the most suitable control actions in order to deal with non-recurrent and non-predictable traffic congestion in a real-time situation. Our intelligent system employs a Fuzzy Neural Networks (FNN) Tool that combines the capabilities of fuzzy reasoning in measuring imprecise and dynamic factors and the capabilities of neural networks in terms of learning processes. In this work we present an effective learning approach with regard to the FNN-Tool, which consists of three stages: initializing the membership functions of both input and output variables by determining their centres and widths using self-organizing algorithms; employing an evolutionary Genetic Algorithm (GA) based learning method to identify the fuzzy rules; tune the derived structure and parameters using the back-propagation learning algorithm. We evaluate experimentally the performance and the prediction capability of this three-stage learning approach using well-known benchmark examples. Experimental results demonstrate the ability of the learning approach to identify all relevant fuzzy rules from the training data. A comparative analysis shows that the proposed learning approach has a higher degree of predictive capability than existing models. We also address the scalability issue of our intelligent traffic control decision support system by using a multi-agent based approach. The large network is divided into sub-networks, each of which has its own associated agent. Finally, our intelligent traffic control decision support system is applied to a number of road traffic case studies using the traffic network in Riyadh, in Saudi Arabia. The results obtained are promising and show that our intelligent traffic control decision support system can provide an effective support for real-time traffic control.
12
Vartak, Aniket Arun. "GAUSS-NEWTON BASED LEARNING FOR FULLY RECURRENT NEURAL NETWORKS." Master's thesis, University of Central Florida, 2004. http://digital.library.ucf.edu/cdm/ref/collection/ETD/id/4429.
Full textAbstract:
The thesis discusses a novel off-line and on-line learning approach for Fully Recurrent Neural Networks (FRNNs). The most popular algorithm for training FRNNs, the Real Time Recurrent Learning (RTRL) algorithm, employs the gradient descent technique for finding the optimum weight vectors in the recurrent neural network. Within the framework of the research presented, a new off-line and on-line variation of RTRL is presented, that is based on the Gauss-Newton method. The method itself is an approximate Newton's method tailored to the specific optimization problem, (non-linear least squares), which aims to speed up the process of FRNN training. The new approach stands as a robust and effective compromise between the original gradient-based RTRL (low computational complexity, slow convergence) and Newton-based variants of RTRL (high computational complexity, fast convergence). By gathering information over time in order to form Gauss-Newton search vectors, the new learning algorithm, GN-RTRL, is capable of converging faster to a better quality solution than the original algorithm. Experimental results reflect these qualities of GN-RTRL, as well as the fact that GN-RTRL may have in practice lower computational cost in comparison, again, to the original RTRL.<br>M.S.<br>Department of Electrical and Computer Engineering<br>Engineering and Computer Science<br>Electrical and Computer Engineering
13
Lustig, Joakim. "Identifying dyslectic gaze pattern : Comparison of methods for identifying dyslectic readers based on eye movement patterns." Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-191233.
Full textAbstract:
Dyslexia affects between 5-17% of all school children, mak-ing it the most common learning disability. It has beenfound to severely affect learning ability in school subjectsas well as limit the choice of further education and occupa-tion. Since research has shown that early intervention andsupport can mitigate the negative effects of dyslexia, it iscrucial that the diagnosis of dyslexia is easily available andaimed at the right children. To make sure children whoare experiencing problems reading and potentially could bedyslectic are investigated for dyslexia an easy access, sys-tematic, and unbiased screening method would be helpful.This thesis therefore investigates the use of machine learn-ing methods to analyze eye movement patterns for dyslexiaclassification.The results showed that it was possible to separatedyslectic from non-dyslectic readers to 83% accuracy, us-ing non-sequential feature based machine learning methods.Equally good results for lower sample frequencies indicatedthat consumer grade eye trackers can be used for the pur-pose. Furthermore a sequential approach using RecurrentNeural Networks was also investigated, reaching an accu-racy of 78%. The thesis is intended to be an introduction to whatmethods could be viable for identifying dyslexia and as aninspiration for researchers aiming to do larger studies in thearea.
14
Kvedaraite, Indre. "Sentiment Analysis of YouTube Public Videos based on their Comments." Thesis, Linnéuniversitetet, Institutionen för datavetenskap och medieteknik (DM), 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:lnu:diva-105754.
Full textAbstract:
With the rise of social media and publicly available data, opinion mining is more accessible than ever. It is valuable for content creators, companies and advertisers to gain insights into what users think and feel. This work examines comments on YouTube videos, and builds a deep learning classifier to automatically determine their sentiment. Four Long Short-Term Memory-based models are trained and evaluated. Experiments are performed to determine which deep learning model performs with the best accuracy, recall, precision, F1 score and ROC curve on a labelled YouTube Comment dataset. The results indicate that a BiLSTM-based model has the overall best performance, with the accuracy of 89%. Furthermore, the four LSTM-based models are evaluated on an IMDB movie review dataset, achieving an average accuracy of 87%, showing that the models can predict the sentiment of different textual data. Finally, a statistical analysis is performed on the YouTube videos, revealing that videos with positive sentiment have a statistically higher number of upvotes and views. However, the number of downvotes is not significantly higher in videos with negative sentiment.
15
Dong, Minjing. "Modelling Skeleton-based Human Dynamics via Retrospection." Thesis, The University of Sydney, 2019. http://hdl.handle.net/2123/21089.
Full textAbstract:
Human motion prediction is one of the key problems in computer vision and robotic vision and has received increasing attention in recent years. The target is to generate the future continuous, realistic human poses given a seed sequence, which can further assist human motion analysis. However, due to the high-uncertainty, it is difficult and challenging to model human dynamics which not only requires spatial information including complicated joint correlations, but also temporal information including periodic properties. Recently, deep recurrent neural networks (RNNs) have achieved impressive success in forecasting human motion with a sequence-to-sequence architecture. However, forecasting in longer time horizons often leads to implausible human poses or converges to mean poses, because of error accumulation and difficulties in keeping track of longer-term information. Based on these observations, in this study, we propose to retrospect human dynamics with attention. A retrospection module is designed upon RNN to regularly retrospect past frames and correct mistakes in time. This significantly improves the memory of RNN and provides sufficient information for the decoder networks to generate longer-term predictions. Moreover, we present a spatial attention module to explore cooperation among joints in performing a particular motion as well as a temporal attention module to exploit the level of importance among observed frames. Residual connections are also included to guarantee the performance of short-term prediction. We evaluate the proposed algorithm on the largest and most challenging Human 3.6M dataset in the field. Experimental results demonstrate the necessity of investigating motion prediction in a self-audit manner and the effectiveness of the proposed algorithm in both short-term and long-term predictions.
16
Melidis, Christos. "Adaptive neural architectures for intuitive robot control." Thesis, University of Plymouth, 2017. http://hdl.handle.net/10026.1/9998.
Full textAbstract:
This thesis puts forward a novel way of control for robotic morphologies. Taking inspiration from Behaviour Based robotics and self-organisation principles, we present an interfacing mechanism, capable of adapting both to the user and the robot, while enabling a paradigm of intuitive control for the user. A transparent mechanism is presented, allowing for a seamless integration of control signals and robot behaviours. Instead of the user adapting to the interface and control paradigm, the proposed architecture allows the user to shape the control motifs in their way of preference, moving away from the cases where the user has to read and understand operation manuals or has to learn to operate a specific device. The seminal idea behind the work presented is the coupling of intuitive human behaviours with the dynamics of a machine in order to control and direct the machine dynamics. Starting from a tabula rasa basis, the architectures presented are able to identify control patterns (behaviours) for any given robotic morphology and successfully merge them with control signals from the user, regardless of the input device used. We provide a deep insight in the advantages of behaviour coupling, investigating the proposed system in detail, providing evidence for and quantifying emergent properties of the models proposed. The structural components of the interface are presented and assessed both individually and as a whole, as are inherent properties of the architectures. The proposed system is examined and tested both in vitro and in vivo, and is shown to work even in cases of complicated environments, as well as, complicated robotic morphologies. As a whole, this paradigm of control is found to highlight the potential for a change in the paradigm of robotic control, and a new level in the taxonomy of human in the loop systems.
17
Shahkarami, Abtin. "Complexity reduction over bi-RNN-based Kerr nonlinearity equalization in dual-polarization fiber-optic communications via a CRNN-based approach." Electronic Thesis or Diss., Institut polytechnique de Paris, 2022. http://www.theses.fr/2022IPPAT034.
Full textAbstract:
Les dégradations dues à la non-linéarité de Kerr dans les fibres optiques limitent les débits d’information des systèmes de communications. Les effets linéaires, tels que la dispersion chromatique et la dispersion modale de polarisation, peuvent être compensés par égalisation linéaire, de mise en oeuvre relativement simple, au niveau du récepteur. A l’inverse, la complexité de calcul des techniques classiques de réduction de la non-linéarité, telles que la rétro-propagation numérique, peut être considérable. Les réseaux neuronaux ont récemment attiré l’attention, dans ce contexte, pour la mise en oeuvre d’égaliseurs non-linéaires à faible complexité. Cette thèse porte sur l’étude des réseaux neuronaux récurrents pour compenser efficacement les dégradations des canaux dans les transmissions à longue distance multiplexés en polarisation. Nous présentons une architecture hybride de réseaux neuronaux récurrents convolutifs (CRNN), comprenant un encodeur basé sur un réseau neuronal convolutif (CNN) suivie d’une couche récurrente travaillant en tandem. L’encodeur basé sur CNN représente efficacement la mémoire de canal à court terme résultant de la dispersion chromatique, tout en faisant passer le signal vers un espace latent avec moins de caractéristiques pertinentes. La couche récurrente suivante est implémentée sous la forme d’un RNN unidirectionnel de type vanille, chargé de capturer les interactions à longue portée négligées par l’encodeur CNN. Nous démontrons que le CRNN proposé atteint la performance des égaliseurs actuels dans la communication par fibre optique, avec une complexité de calcul significativement plus faible selon le modèle du système. Enfin, le compromis performance-complexité est établi pour un certain nombre de modèles, y compris les réseaux neuronaux multicouches entièrement connectés, les CNN, les réseaux neuronaux récurrents bidirectionnels, les réseaux long short-term memory bidirectionnels (bi-LSTM), les réseaux gated recurrent units bidirectionnels, les modèles bi-LSTM convolutifs et le modèle hybride proposé<br>The impairments arising from the Kerr nonlinearity in optical fibers limit the achievable information rates in fiber-optic communication. Unlike linear effects, such as chromatic dispersion and polarization-mode dispersion, which can be compensated via relatively simple linear equalization at the receiver, the computational complexity of the conventional nonlinearity mitigation techniques, such as the digital backpropagation, can be substantial. Neural networks have recently attracted attention, in this context, for low-complexity nonlinearity mitigation in fiber-optic communications. This Ph.D. dissertation deals with investigating the recurrent neural networks to efficiently compensate for the nonlinear channel impairments in dual-polarization long-haul fiber-optic transmission. We present a hybrid convolutional recurrent neural network (CRNN) architecture, comprising a convolutional neural network (CNN) -based encoder followed by a recurrent layer working in tandem. The CNN-based encoder represents the shortterm channel memory arising from the chromatic dispersion efficiently, while transitioning the signal to a latent space with fewer relevant features. The subsequent recurrent layer is implemented in the form of a unidirectional vanilla RNN, responsible for capturing the long-range interactions neglected by the CNN encoder. We demonstrate that the proposed CRNN achieves the performance of the state-of-theart equalizers in optical fiber communication, with significantly lower computational complexity depending on the system model. Finally, the performance complexity trade-off is established for a number of models, including multi-layer fully-connected neural networks, CNNs, bidirectional recurrent neural networks, bidirectional long short-term memory (bi-LSTM), bidirectional gated recurrent units, convolutional bi-LSTM models, and the suggested hybrid model
18
AbuRa'ed, Ahmed Ghassan Tawfiq. "Automatic generation of descriptive related work reports." Doctoral thesis, Universitat Pompeu Fabra, 2020. http://hdl.handle.net/10803/669975.
Full textAbstract:
A related work report is a section in a research paper which integrates key information from a list of related scientific papers providing context to the work being presented. Related work reports can either be descriptive or integrative. Integrative related work reports provide a high-level overview and critique of the scientific papers by comparing them with each other, providing fewer details of individual studies. Descriptive related work reports, instead, provide more in-depth information about each mentioned study providing information such as methods and results of the cited works.
In order to write a related work report, scientist have to identify, condense/summarize, and combine relevant information from different scientific papers. However, such task is complicated due to the available volume of scientific papers. In this context, the automatic generation of related work reports appears to be an important problem to tackle. The automatic generation of related work reports can be considered as an instance of the multi-document summarization problem where, given a list of scientific papers, the main objective is to automatically summarize those scientific papers and generate related work reports.
In order to study the problem of related work generation, we have developed a manually annotated, machine readable data-set of related work sections, cited papers (e.g. references) and sentences, together with an additional layer of papers citing the references. We have also investigated the relation between a citation context in a citing paper and the scientific paper it is citing so as to properly model cross-document relations and inform our summarization approach. Moreover, we have also investigated the identification of explicit and implicit citations to a given scientific paper which is an important task in several scientific text mining activities such as citation purpose identification, scientific opinion mining, and scientific summarization.
We present both extractive and abstractive methods to summarize a list of scientific papers by utilizing their citation network. The extractive approach follows three stages: scoring the sentences of the scientific papers based on their citation network, selecting sentences from each scientific paper to be mentioned in the related work report, and generating an organized related work report by grouping the sentences of the scientific papers that belong to the same topic together. On the other hand, the abstractive approach attempts to generate citation sentences to be included in a related work report, taking advantage of current sequence-to-sequence neural architectures and resources that we have created specifically for this task.
The thesis also presents and discusses automatic and manual evaluation of the generated related work reports showing the viability of the proposed approaches.<br>La sección de trabajos relacionados de un artículo científico resume e integra información clave de una lista de documentos científicos relacionados con el trabajo que se presenta. Para redactar esta sección del artículo científico el autor debe identificar, condensar/resumir y combinar información relevante de diferentes artículos. Esta tarea es complicada debido al gran volumen disponible de artículos científicos. En este contexto, la generación automática de tales secciones es un problema importante a abordar. La generación automática de secciones de trabajo relacionados puede ser considerada como una instancia del problema de resumen de documentos múltiples donde, dada una lista de documentos científicos, el objetivo es resumir automáticamente esos documentos científicos y generar la sección de trabajos relacionados.
Para estudiar este problema, hemos creado un corpus de secciones de trabajos relacionados anotado manualmente y procesado automáticamente. Asimismo, hemos investigado la relación entre las citaciones y el artículo científico que se cita para modelar adecuadamente las relaciones entre documentos y, así, informar nuestro método de resumen automático. Además, hemos investigado la identificación de citaciones implícitas a un artículo científico dado que es una tarea importante en varias actividades de minería de textos científicos.
Presentamos métodos extractivos y abstractivos para resumir una lista de artículos científicos utilizando su red de citaciones. El enfoque extractivo sigue tres etapas: cálculo de la relevancia las oraciones de cada artículo en función de la red de citaciones, selección de oraciones de cada artículo científico para integrarlas en el resumen y generación de la sección de trabajos relacionados agrupando las oraciones por tema. Por otro lado, el enfoque abstractivo intenta generar citaciones para incluirlas en un resumen utilizando redes neuronales y recursos que hemos creado específicamente para esta tarea.
La tesis también presenta y discute la evaluación automática y manual de los resúmenes generados automáticamente, demostrando la viabilidad de los enfoques propuestos.<br>Una secció d’antecedents o estat de l’art d’un articulo científic resumeix la informació clau d'una llista de documents científics relacionats amb el treball que es presenta. Per a redactar aquesta secció de l’article científic l’autor ha d’identificar, condensar / resumir i combinar informació rellevant de diferents articles. Aquesta activitat és complicada per causa del gran volum disponible d’articles científics. En aquest context, la generació automàtica d’aquestes seccions és un problema important a abordar. La generació automàtica d’antecedents o d’estat de l’art pot considerar-se com una instància del problema de resum de documents.
Per estudiar aquest problema, es va crear un corpus de seccions d’estat de l’art d’articles científics manualment anotat i processat automàticament. Així mateix, es va investigar la relació entre citacions i l’article científic que es cita per modelar adequadament les relacions entre documents i, així, informar el nostre mètode de resum automàtic. A més, es va investigar la identificació de citacions implícites a un article científic que és un problema important en diverses activitats de mineria de textos científics.
Presentem mètodes extractius i abstractius per resumir una llista d'articles científics utilitzant el conjunt de citacions de cada article. L’enfoc extractiu segueix tres etapes: càlcul de la rellevància de les oracions de cada article en funció de les seves citacions, selecció d’oracions de cada article científic per a integrar-les en el resum i generació de la secció de treballs relacionats agrupant les oracions per tema. Per un altre costat, l’enfoc abstractiu implementa la generació de citacions per a incloure-les en un resum que utilitza xarxes neuronals i recursos que hem creat específicament per a aquest tasca.
La tesi també presenta i discuteix l'avaluació automàtica i el manual dels resums generats automàticament, demostrant la viabilitat dels mètodes proposats.
19
Arcolezi, Héber Hwang. "A novel robust and intelligent control based approach for human lower limb rehabilitation via neuromuscular electrical stimulation /." Ilha Solteira, 2019. http://hdl.handle.net/11449/190755.
Full textAbstract:
Orientador: Aparecido Augusto de Carvalho<br>Abstract: In the last few years, several studies have been carried out showing that neuromuscular electrical stimulation (NMES) can produce good therapeutic results in patients with spinal cord injury (SCI). This research introduces a new robust and intelligent control-based methodology for human lower limb rehabilitation via NMES using a continuous-time control technique named robust integral of the sign of the error (RISE). Although in the literature the RISE controller has shown good results without any fine-tuning method, a trial and error approach would quickly lead to muscle fatigue in SCI patients. Therefore, it was shown in this study that the control performance for robustly tracking a reference signal can be improved through the proposed approach by providing an intelligent tuning for each voluntary. Simulation results with a mathematical model and eight identified subjects from the literature are provided, and real experiments are performed with seven healthy and two paraplegic subjects. Besides, this research introduces the application of deep and dynamic neural networks namely the multilayer perceptron, a simple recurrent neural network, and the Long Short-Term memory architecture, to identify the nonlinear and time-varying relationship between the supplied NMES and achieved angular position. Identification results indicate good fitting to data and very low mean square error using few data for training, proving to be very prospective methods for proposing control-oriented ... (Complete abstract click electronic access below)<br>Resumo: Nos últimos anos, vários estudos foram realizados mostrando que a estimulação elétrica neuromuscular (EENM) pode produzir bons resultados terapêuticos em pacientes com lesão medular (LM). Esta pesquisa introduz uma nova metodologia robusta e inteligente baseada em controle para a reabilitação de membros inferiores humanos via EENM usando uma técnica de controle de tempo contínuo chamada robust integral of the sign of the error (RISE). Embora na literatura o controlador RISE tem demonstrado bons resultados sem qualquer método de ajuste fino, uma abordagem de tentativa e erro poderia levar rapidamente à fadiga muscular em pacientes com LM. Portanto, foi mostrado nesse estudo que o desempenho do controle para rastrear com robustez um sinal de referência pode ser melhorado através da abordagem proposta, fornecendo um ajuste inteligente para cada voluntário. Resultados de simulação com um modelo matemático e oito sujeitos identificados da literatura são fornecidos, e experimentos reais são feitos com sete indivíduos saudáveis e dois paraplégicos. Além disso, esta pesquisa introduz a aplicação de redes neurais profundas e dinâmicas, especificamente o perceptron multicamadas, uma rede neural recorrente simples e a arquitetura Long Short-Term Memory, para identificar a relação não-linear e variante no tempo entre a EENM fornecida e a posição angular alcançada. Os resultados de identificação indicam boa adaptação aos dados e erro quadrático médio muito baixo usando poucos dados para... (Resumo completo, clicar acesso eletrônico abaixo)<br>Mestre
20
Kleman, Björn, and Henrik Lindgren. "Evaluation of model-based fault diagnosis combining physical insights and neural networks applied to an exhaust gas treatment system case study." Thesis, Linköpings universitet, Fordonssystem, 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-176650.
Full textAbstract:
Fault diagnosis can be used to early detect faults in a technical system, which means that workshop service can be planned before a component is fully degraded. Fault diagnosis helps with avoiding downtime, accidents and can be used to reduce emissions for certain applications. Traditionally, however, diagnosis systems have been designed using ad hoc methods and a lot of system knowledge. Model-based diagnosis is a systematic way of designing diagnosis systems that is modular and offers high performance. A model-based diagnosis system can be designed by making use of mathematical models that are otherwise used for simulation and control applications. A downside of model-based diagnosis is the modeling effort needed when no accurate models are available, which can take a large amount of time. This has motivated the use of data-driven diagnosis. Data-driven methods do not require as much system knowledge and modeling effort though they require large amounts of data and data from faults that can be hard to gather. Hybrid fault diagnosis methods combining models and training data can take advantage of both approaches decreasing the amount of time needed for modeling and does not require data from faults. In this thesis work a combined data-driven and model-based fault diagnosis system has been developed and evaluated for the exhaust treatment system in a heavy-duty diesel engine truck. The diagnosis system combines physical insights and neural networks to detect and isolate faults for the exhaust treatment system. This diagnosis system is compared with another system developed during this thesis using only model-based methods. Experiments have been done by using data from a heavy-duty truck from Scania. The results show the effectiveness of both methods in an industrial setting. It is shown how model-based approaches can be used to improve diagnostic performance. The hybrid method is showed to be an efficient way of developing a diagnosis system. Some downsides are highlighted such as the performance of the system developed using data-driven and model-based methods depending on the quality of the training data. Future work regarding the modularity and transferability of the hybrid method can be done for further evaluation.
21
Xu, Wenduan. "Structured learning with inexact search : advances in shift-reduce CCG parsing." Thesis, University of Cambridge, 2017. https://www.repository.cam.ac.uk/handle/1810/270026.
Full textAbstract:
Statistical shift-reduce parsing involves the interplay of representation learning, structured learning, and inexact search. This dissertation considers approaches that tightly integrate these three elements and explores three novel models for shift-reduce CCG parsing. First, I develop a dependency model, in which the selection of shift-reduce action sequences producing a dependency structure is treated as a hidden variable; the key components of the model are a dependency oracle and a learning algorithm that integrates the dependency oracle, the structured perceptron, and beam search. Second, I present expected F-measure training and show how to derive a globally normalized RNN model, in which beam search is naturally incorporated and used in conjunction with the objective to learn shift-reduce action sequences optimized for the final evaluation metric. Finally, I describe an LSTM model that is able to construct parser state representations incrementally by following the shift-reduce syntactic derivation process; I show expected F-measure training, which is agnostic to the underlying neural network, can be applied in this setting to obtain globally normalized greedy and beam-search LSTM shift-reduce parsers.
22
LIN, CHENG-YANG, and 林政陽. "Recurrent Neural Network-based Microphone Howling Suppression." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/hd839v.
Full textAbstract:
碩士<br>國立臺北科技大學<br>電子工程系<br>107<br>When using the karaoke system to sing, it is often too close the microphone and power of the amplified speaker is too large, causing a positive feedback and howling making the singer and the listener to be uncomfortable. Generally, to solve the microphone howling, often using a frequency shift to interrupt the resonance, or using a band-stop filter to remedy afterwards. But both may cause sound quality damage. Therefore, we want to use the adaptive feedback cancellation algorithm. Using the input source of the amplified speaker as the reference signal to automatically estimate the feedback signals that may record in different signal-to-noise. And eliminate the signal gain before howling occurs directly from the source. Based on the above ideas, in this paper, the howling elimination algorithm of normalized least mean square (NLMS) is realized, especially considering the nonlinear distortion of the sound amplification system, and the advanced algorithm based on recurrent neural network (RNN) is proposed. And in the experiment, test the time-domain or frequency-domain processing separately, and use NLMS or RNN, a total of four different combinations, the convergence speed and computational demand of different algorithms under different temperament and different environmental spatial response situations and howling suppression effect. The experimental results show that: (1) the convergence in the time domain is faster, (2) Stable effect in the frequency domain (3) Time domain RNN is best at eliminating effects, but there are too large calculations.
23
Tsai, Yao-Cheng, and 蔡曜丞. "Acoustic Echo Cancellation Based on Recurrent Neural Network." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/jgk3ea.
Full textAbstract:
碩士<br>國立中央大學<br>通訊工程學系<br>107<br>Acoustic echo cancellation is a common problem in speech and signal processing until now. Application scenarios such as telephone conference, hands-free handsets and mobile communications. In the past we used adaptive filters to deal with acoustic echo cancellation, and today we can use deep learning to solve complex problems in acoustic echo cancellation.
The method proposed in this work is to consider acoustic echo cancellation as a problem of speech separation, instead of the traditional adaptive filter to estimate acoustic echo. And use the recurrent neural network architecture in deep learning to train the model. Since the recurrent neural network has a good ability to simulate time-varying functions, it can play a role in solving the problem of acoustic echo cancellation. We train a bidirectional long short-term memory network and a bidirectional gated recurrent unit. Features are extracted from single-talk speech and double-talk speech. Adjust weights to control the ratio between double-talk speech and single-talk speech, and estimate the ideal ratio mask. This way to separate the signal, in order to achieve the purpose of removing the echo. The experimental results show that the method has good effect in echo cancellation.
24
Hu, Hsiao-Chun, and 胡筱君. "Recurrent Neural Network based Collaborative Filtering Recommender System." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/ytva33.
Full textAbstract:
碩士<br>國立臺灣科技大學<br>資訊工程系<br>107<br>As the rapid development of e-commerce, Collaborative Filtering Recommender System has been widely applied to major network platforms. Predict customers’ preferences accurately through recommender system could solve the problem of information overload for users and reinforce their dependence on the network platform. Since the recommender system based on collaborative filtering has the ability to recommend products that are abstract or difficult to describe in words, research related to collaborative filtering has attracted more and more attention. In this paper, we propose a deep learning model framework for collaborative filtering recommender system. We use Recurrent Neural Network as the most important part of this framework which makes our model have the ability to consider the timestamp of implicit feedbacks from each user. This ability then significantly improve the performance of our models when making personalization item recommendations. In addition, we also propose a training data format for Recurrent Neural Network. This format makes our recommender system became the first Recurrent Neural Network model that can consider both positive and negative implicit feedback instance during the training process. Through conducted experiments on the two real-world datasets, MovieLens-1m and Pinterest, we verify that our model can finish the training process during a shorter time and have better recommendation performance than the current deep learning based Collaborative Filtering model.
25
CHEN, SHEN-CHI, and 陳順麒. "On the Recurrent Neural Network Based Intrusion Detection System." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/75tb39.
Full textAbstract:
碩士<br>逢甲大學<br>資訊工程學系<br>107<br>With the advancement of modern science and technology, numerous applications of the Internet of Things are developing faster and faster. Smart grid is one of the examples which provides full communication, monitor, and control abilities to the components in the power systems in order to meet the increasing demands of reliable energy. In such systems, many components can be monitored and controlled remotely. As a result, they could be vulnerable to malicious cyber-attacks if there exist exploitable loopholes. In the power system, the disturbances caused by cyber-attacks are mixed with those caused by natural events. It is crucial for the intrusion detection systems in the smart grid to classify the types of disturbances and pinpoint the attacks with high accuracy.
The amount of information in a smart grid system is much larger than before, and the amount of computation of the big data increases accordingly. Many analyzing techniques have been proposed to extract useful information in these data and deep learning is one of them. It can be applied to “learn” a model from a large set of training data and classify unknown events from subsequent data. In this paper, we apply the methods of recurrent neural network (RNN) algorithm as well as two other variants to train models for intrusion detection in smart grid. Our experiment results showed that RNN can achieves high accuracy and precision on a set of real data collected from an experimental power system network.
26
Wu, Ting-Xuan, and 吳亭萱. "Image Compression Using LSTM/GRU Based Recurrent Convolution Neural Network." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/bqk75g.
Full textAbstract:
碩士<br>國立臺灣科技大學<br>電機工程系<br>106<br>Performances of Image/video encoding methods, such as robustness to error attack and compression ratio, are important for multimedia communication applications. The JPEG image compression standard adopts block-based discrete cosine transform (DCT) and the JPEG2000 utilize wavelet transform (WT) to provide multi-resolution compression. Both standards are efficient for current multimedia communication standards. In this research, we study how to utilize deep learning methods to compress image signals, which can provide comparable performances with DCT-based JPEG and WT-based JPEG2000.
We proposed an auto encoder architecture for image compression based on a multi-layer recurrent convolutional neural network that comprises an encoder and a decoder sub models. For network nodes, we use long short-term memory network (LSTM) or GRU (Gated Recurrent Units) to enable efficient information delivery. In the training phase, both encoding and decoding procedures are trained together to reserve the most image feature information during compression and decompression. . In the testing phase, it executes encoding and decoding procedures separately to verify performances. Experiments showed that when BPP > 0.8, the PSNR and MS-SSIM (Multiscale Structure Similarity) performances of the proposed methods are better than those of JPEG. For high compression BPP < 0.8, the proposed method outperforms JPEG and others in MS-SSIM, which demonstrates better visual perception performance, i.e., sharper edge, rich textures and fewer artifacts.
27
Chou, Yung-Chieh, and 周詠捷. "Automatic Term Explanation based on Topic-regularized Recurrent Neural Network." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/558543.
Full textAbstract:
碩士<br>國立中山大學<br>資訊管理學系研究所<br>106<br>In this study, we propose a topic-regularized Recurrent Neural Network(RNN)-based model designed to explain given terms. RNN-based models usually generate text results that have correct syntax but lack coherence, whereas topic models produce several topics consisting of coherent keywords. Here we consider combining them into a new model that takes advantages of both. In our experiment, we trained Long Short-Term Memory (LSTM) models on selected articles that mention given terms, applying nonsmooth nonnegative matrix factorization(nsNMF) on document-term matrix to obtain contextual biases. Our empirical results showed that topic-regularizing LSTM outperforms original models while generating readable sentences. Additionally, topic-regularized LSTM could adopt different topics to generate description about subtle but important aspects of a certain field, which is usually not captured by original LSTM.
28
Hong, Chun-Ming, and 洪俊銘. "Permanent Magnet Linear Synchronous Motor Based on Recurrent Neural Network Controller." Thesis, 2000. http://ndltd.ncl.edu.tw/handle/77286220803715560477.
Full textAbstract:
碩士<br>中原大學<br>電機工程學系<br>88<br>The purpose of this thesis is to control permanent magnet linear synchronous motor (PMLSM) drive system to track perodic commands using the robust control system based on recurrent neural network (RNN). First, an integral-proportional (IP) position controller is introduced to control the mover position of the PMLSM. The IP position controller is designed according to the estimated mover parameters to match the time-domain command tracking specifications. Then, a disturbance observer is implemented and the observed disturbance force is fed forward to increase the robustness of the PMLSM drive system. Moreover, to increase the control performance of the PMLSM drive system under the occurrence of parameter variations and external disturbance, a RNN compensator is proposed to reduce the influence of parameter variations and external disturbances of the PMLSM drive system. Furthermore, to increase the robustness of the PMLSM drive system for different periodic command inputs, a RNN controller is proposed to control the mover position of the PMLSM. Finally, a hybrid supervisory control system, which combines a supervisory control system and an intelligent control system using recurrent fuzzy neural network (RFNN), is proposed to control the mover of the PMLSM for periodic motion. The effectiveness of the proposed control schemes is demonstrated by some simulated and experimental results.
29
Tsung-ChiehWen and 溫淙傑. "Implementation of Text Classification Model Based on Recurrent Convolutional Neural Network." Thesis, 2017. http://ndltd.ncl.edu.tw/handle/qq5p25.
Full textAbstract:
碩士<br>國立成功大學<br>工程科學系碩士在職專班<br>105<br>In a variety of community media types of network platform have on-line operation, and the popularity of smart phones, these changes have changed the way people use the web. Over the past years, users only can search and get the information from the website, but now the user can be an information provider. Many Internet users began to be willing and keen to share their views out, so there is a lot of text data on the Internet. Ten years ago people have heard this sentence:「This is an era of information explosion」, and now because everyone can be the provider of messages, as compared to a decade ago, the current amount of information on the Internet is more larger.
These content generated by the user often contains some opinion, evaluation and other information, and these messages can often be converted into valuable information, and this information can be used by individuals or corporate groups. But the text on the Internet is too much, cannot be man-made to collect and analyze. So how to use the machine to help users analyze these texts is one of the important topics in the field of information capture in recent years.
In this research, we implemented a deep learning network architecture, which combined with the architecture of convolutional neural network and the architecture of recurrent neural network. And use to complete the goal of text classification with the pre-trained word vectors. The difficulty is that the word vector should be used as a static lookup table without updating, but the network still can ignore the noise which caused by missing words to complete the task. The experimental results show that the accuracy of this study is consistent with the accuracy of other studies, proved the feasibility of this architecture. And has the following advantages: 1. The accuracy rate of this architecture is higher than that of recurrent neural network, 2. Compared with the convolution neural network, the accuracy results are more stable, 3. Use less epoch to get stable results. But the shortcoming of this research architecture is that training time is too long.
30
Wang, Wei-Di, and 王偉地. "The Study of Option Pricing——Based on the Recurrent Neural Network." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/54y3f5.
Full textAbstract:
碩士<br>國立臺灣大學<br>財務金融學研究所<br>106<br>Previous studies have shown that neural network models which are trained by historical data can be used to price options. Because of no assumption needed, neural network has better performance than those traditional parameter models such as Black-Scholes in pricing. Those neural network models used in the past have two characteristics. One is that those models have only several hidden layers, and the other is that only the volatility is used as a variable to describe the fluctuation of the underlying security when designing the input variables.
With the development of artificial intelligence-related technologies in recent years, some neural network models based on deep learning are applied to many fields of society. In this paper, we try to use the recurrent neural network to study the pricing of options, and directly put the time series data of the underlying security as input variables into the model, hoping to hoping to get a neural network model which has better performance than the formers do.
Shanghai Stock Exchange 50ETF option is selected as the object of this study. After the adjustment and selection of neural network model structure, this paper selects a neural network model with one hidden layer, two recurrent neural network models and a Black-Scholes model. We use these four models to price options and compare the performance in pricing accuracy. The final results show that in terms of the pricing accuracy, the recurrent neural network model is significantly superior to the Black-Scholes model and the neural network model with one hidden layer during the selected data samples.
31
Huang, Po-Cheng, and 黃柏程. "A Hybrid Facial Expression Recognition System based on Recurrent Neural Network." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/54ugqm.
Full textAbstract:
碩士<br>國立臺灣科技大學<br>電機工程系<br>107<br>Facial expression recognition (FER) based on facial features is an important and challenging problem for automatic inspection of surveillance videos. In this thesis, a hybrid facial expression recognition system is proposed based on facial features, and a total of six facial expressions can be recognized. For the pre-processing, the Random Forest classifier is applied to track the facial landmark points, which is utilized for face alignment. In the feature extraction, with the advance of deep learning technology, we introduced the hybrid RNN technique by incorporating with deep learning to extract robust features. In addition, the geometrical features and the facial action unit based on the movement of the facial feature point are also considered. As the results, we evaluate the proposed method on the two database, CK+ and Oulu-CASIA, and compare with previous works. Though there are uncontrolled factors in the videos, such as lighting and head posture, the proposed method can achieve superior performance than former schemes. Thus, the proposed method has considerable potential to be applied in the practical applications.
32
Jan-ShianLiu and 劉展憲. "Study on Hand Tracking and Dynamic Hand Gesture Recognition based on Convolutional Neural Network and Recurrent Neural Network." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/n5bkf6.
Full text
33
Tzeng, Mao-Sheng, and 曾楙升. "Speed Sensorless Induction Spindle Motor Based on Recurrent Fuzzy Neural Network Controller." Thesis, 2001. http://ndltd.ncl.edu.tw/handle/62507877979435704194.
Full textAbstract:
碩士<br>中原大學<br>電機工程研究所<br>89<br>The purpose of this thesis is to develop a high frequency drive system with advanced rotor speed estimation and control algorithms for the three-phase induction spindle motor. First, a new type synchronous pulse width modulation (PWM) techniques with 1KHz switching frequency is developed to reduce the harmonic components of the output voltages and currents of the high-frequency inverter. Moreover, the dead-time compensation technique using FPGA is developed to increase the voltage utilization of the inverter. The high-frequency inverter is implemented using IGBT switching components. Then a personal computer (PC) is adopted to control the motor drive system with voltage/frequency ( ) constant control. On the other hand, since the tachometer and encoder are impossible to couple to the rotor of the spindle motor for rotor speed higher than few ten thousands rpm, the rotor speed and dynamic can not be controlled precisely. Therefore, the development of advanced rotor speed estimator is necessary. Furthermore, the speed controller is developed using recurrent fuzzy neural network (RFNN). The theoretical basis of the proposed estimator and controller are derived in detail, and the effectiveness of the proposed spindle motor drive system is confirmed by simulation and experimental results.
34
Wu, Jian-Huei, and 吳建輝. "Hammerstein Recurrent Neural Network Based Nonlinear Model Predictive Control and Its Applications." Thesis, 2007. http://ndltd.ncl.edu.tw/handle/77802265139771223116.
Full textAbstract:
碩士<br>國立成功大學<br>電機工程學系碩博士班<br>95<br>This thesis presents a nonlinear model predictive control (NMPC) scheme based on a Hammerstein recurrent neural network (HRNN) for controlling unknown systems. The unknown system is first modeled by the HRNN that consists of a static nonlinear subsystem cascaded by a dynamic linear subsystem. The HRNN is capable of transferring an unknown dynamic system into a state-space representation. An effective construction algorithm, which integrates the methods of order determination, parameter initialization and performance optimization, is utilized to construct a parsimonious HRNN with a satisfactory performance. The philosophy of our NMPC scheme system is to establish a nonlinearity eliminator that functions as the inverse of the static nonlinear model to remove the nonlinear behavior of the unknown system. If the system modeling and the inverse of the nonlinear model are accurate, the compound model, the unknown system cascaded with the nonlinearity eliminator, will behave like the linear dynamic model. Hence, the theories of linear model predictive controller design can be applied directly to achieve good control performance. Computer simulations on nonlinear system control problems have successfully validated the effectiveness of the proposed control scheme.
35
Wang, Mu-Fan, and 王牧凡. "SPENT : Similarity-based POI Embedding and recurrent Neural network with Temporal influence." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/w58qu2.
Full textAbstract:
碩士<br>國立交通大學<br>資訊科學與工程研究所<br>106<br>Widespread smartphone use has changed human lifestyle gradually. Nowadays, people may take pictures, leave comments or score it in location-based social networks(LBSNs) to share their life experience. On the other hand, people can browse other people’s check-ins
or get recommendations from LBSN service to determine where to visit. In recent years, successive poi recommendation has got more and more attention, which considers current location and the correlation between POIs in addition. For example, people may visit a museum then go to a nearby restaurant which has been visited by many tourists or locals. Through profiling users and POIs by check-in sequences, we can enhance the quality of recommendation so as to figure out users’ preference in locations. In this paper, we propose the SPENT model, based on a variant of Word2Vec technique and a recurrent
neural network that jointly learn the complex time-aware influence. We demonstrate our model on two well-known datasets, Gowalla and Foursquare, and show the effectiveness of each factor.
36
Wang, Chun-Hao, and 王君豪. "Sound Event Detection Based on Partitioned Autoencoder and Convolutional Recurrent Neural Network." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/7rna2b.
Full textAbstract:
碩士<br>國立清華大學<br>電機工程學系<br>107<br>In this thesis, a noise reduction process and a sound event detection (SED) system are used in detecting DCASE2017 TUT Sound Events 2017 dataset [1] which contains six sound events in a total of 32 audio recordings (24 audio recordings in the development set and 8 audio recordings in the evaluation set). It is a polyphonic task and the trained SED model have to detect the sound events with their onset time and offset time. The purpose of the noise reduction is to observe whether it is helpful in the training process of the sound event detection task. In this thesis, a partitioned autoencoder [2] is adopted for noise reduction. In the sound event detection part, a convolutional recurrent neural network (CRNN) [3][4] which won the first prize in the task of "sound event detection in real life" in DCASE2017 is adopted. The original log mel-band energies, the denoised log mel-band energies, and the augmented log mel-band energies which combine both of above are the input features of the CRNN. From the training results, it reveals that the SED models trained with the denoised features have better performance in some sound events by showing the lower medians of the testing error rates or showing the better distribution of the testing error rates. Furthermore, the final performance of the error rates reveals that the models trained with the denoised features can achieve the best testing error rate of 0.622 in the development set and the best testing error rate of 0.744 in the evaluation set. The testing error rate could be improved further if choosing the best model of each class across 3 kinds of features.
37
Hsieh, Yi-Hsin, and 謝沂歆. "Predicting customer behavior with Recurrent Neural Network based on trust and loyalty." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/99c497.
Full textAbstract:
碩士<br>國立中央大學<br>企業管理學系<br>106<br>In the age of rising awareness in health, preventative medication has brought many attentions. In addition, this concept has resulted in the demand of nutrition and food supplements. The sales of these products highly depend on customer trust and loyalty. Therefore, trust and loyalty may be used to predict repurchase behavior.
In the past studies, trust and loyalty has to be analyzed with data collected with questionnaires. The common problem with this approach is low returning rate and less reliable data. This study is collect data through call records and transaction records of the nutrition and health industry through telemarketing channel. The data are used to quantify trust and loyalty. With these values, Recurrent Neural Networks(RNN) method is utilized to predict customer's future repurchase behavior. The results show that the accuracy can reach 70%, which is higher than the accuracy derived from SVM. Companies can use these indicators (trust and loyalty) as KPIs to adjust corporate sales methods and improve the interactions with customers.
38
Chuang, Ying-Wei, and 莊英瑋. "Driver Behavior Recognition based on Multiple Depth Cameras using Recurrent Neural Network." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/dpzy83.
Full textAbstract:
碩士<br>國立中央大學<br>通訊工程學系<br>106<br>This thesis is aimed at in-car driver behavior recognition. One of the purpose is for the safe drive, because it would be dangerous that driver doesn’t concentrate when driving. The other is the application for the In-car entertainment. We propose a multi-view driver behavior recognition system (MDBR system). The pointcloud is captured from different views, and we manage to preprocess the original data by rotation, calibration, merging and sampling. Then, we use the Long short-term memory (LSTM) network, a type of recurrent neural network, as classifier. The dataset we used is VAP multi-view driver behavior dataset. This dataset is we proposed, and contain 10 driver behavior. Using multi-view data can effectively reduce the influence of the occlusion problem. The recognition accuracy of MDBR system have good performance.
39
Mou, Ting-Chen, and 牟庭辰. "Deadlift Recognition and Application based on Multiple Modalities using Recurrent Neural Network." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/9we9pm.
Full textAbstract:
碩士<br>國立中央大學<br>通訊工程學系<br>107<br>With the rise of artificial intelligence and the improvement of computer hardware performance in recent years, the human action recognition (HAR) has gradually been popular, especially in Computer Vision and Pattern Recognition. The application has been widely developed in various field. For example, games, tracking and surveillance systems, smart environment, medical field etc.
This thesis is aimed at fitness behavior recognition, taking deadlift as an example. This exercise is a multi-joint movement, however beginners usually cause injury due to their incorrect concept and wrong posture. One of the purpose is for giving users some professional advice about fitness. The other is the system for the gym can share the work of the fitness instructor, so coaches can focus on the more professional teaching content to their students.
We propose a multi-modality deadlift recognition and application system, including Kinect camera and inertial sensors (x-OSC). Using Time-Variant Skeleton Vector Projection method and feature concatenated method before we feed features to our network. Then, we use the Long short-term memory (LSTM) network, a type of recurrent neural network, as classifier. The dataset we used is VAP multi-modalities fitness behavior dataset. This dataset is we proposed, and contain 6 fitness behavior. Using multi-modality data can achieve a good recognition accuracy effectively and the applications of our system can also analyze uses results effectively.
40
CHUANG, YI-TING, and 莊宜庭. "A PM2.5 Prediction Model Based on Deep Learning with Recurrent Neural Network." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/sfu623.
Full textAbstract:
碩士<br>東海大學<br>資訊管理學系<br>107<br>In recent years, many studies have verified that air pollution will seriously affect human health. In addition, the media reported many issues concerning air pollution, so people have begun to pay attention to its existence. This study analyzes the data of the Environmental Protection Administration air quality immediate pollution indicators in 2018. Five methods are used to deal with the missing values. The main correlation variables affecting the PM25 concentration are identified by principal component analysis and correlation coefficients (single factor: PM10, SO2, NOX, NO2, CO, two-factor: NOX+NO2+CO, SO2+PM10), and the Long-Short Term Memory Model (LSTM) of the Recurrent Neural Network (RNN) was used to model the PM25 concentration model for the next 8 hours. According to the research results, most of the errors between the predicted and true values of Fengyuan Station fall within the reasonable range of MAPE (0.2~0.5). In addition, the best way to deal with the missing value is linear interpolation.
41
Pe#westeur050#a, Anthony Spence, and 施東尼. "A recurrent neural network travel route recommendation system based on geotagged photos." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/mn5q5n.
Full textAbstract:
碩士<br>元智大學<br>工業工程與管理學系<br>107<br>Traveling is an important component of every human life. The tourism industry has experienced a steady growth over the years. Personalization has been found to be one of the key factors of growth in the recent years. However, the large volume of information available has hindered travelers’ ability to make travel plans in a fast and easy fashion. In order to address this problem a travel route recommendation system is proposed. This recommender was developed by inferring trip information obtained from metadata of photos posted by travelers on FLICKR. Locations visited were extracted by means of geospatial clustering. The types of travelers were inferred by executing topic modelling and clustering the resulting vectors. Total trip duration was calculated as the sum of the time spent at each location plus the travel time between locations. The recommender itself is built on a recurrent neural network. Based on the sequences learned and the particular user class to which the sequences belong to, the recommender predicts the next most likely location to be visited, while the trip duration acts as a constraint on the length of the trip. The experiments showed that user classes have a high degree of influence in the final result of the recommendation. Also, constraining the recommendation by using the trip duration showed to be an effective way of providing recommendations bounded in the user’s context of time available. The recommender not only does not repeat locations, but also provides logical trips that are shorter in terms of kilometers in comparison to a Markov Model recommendation system.
42
Zhang, Bin, and 張斌. "A Multiple Time Series-based Recurrent Neural Network for Short-Term Load Forecasting." Thesis, 2017. http://ndltd.ncl.edu.tw/handle/v74yf9.
Full textAbstract:
碩士<br>元智大學<br>資訊工程學系<br>105<br>Electricity, is an indispensable resource in daily life and industrial production. It is related to the national economic construction and the steady development of various industries. However, the production and consumption of electricity resources should be synchronization due to non-mass storage. It is important that the accurate forecasting of load for both the utility and the energy sector, especially for short-term load forecasting(STLF). In the past research, taking main the climatic conditions, historical power load changes, and other factors into account, while the dependence analysis between the various factors become very difficult when we deal with the more complex and diversified data structures and rapid changes, result to the forecast accuracy and stability not be guaranteed. In this paper, we propose a recurrent neural network model based on the deep learning framework(MTS-RNN) to analyze the short-term load forecasting. We first propose the concept of multiple time series, according to the different time-steps, we divide the historical power load into continuous sequences like short-term series and long short-term series, as well as discrete sequences like cycle series and cross long short-term series with jumping time-steps. Then, a deep recurrent neural network model is used to train the single and multiple time series, and analyze it. Experiments show that the combined model has high forecasting performance presented in this paper and superior to other methods in the same data set. The best mean absolute percent error(MAPE) is 0.71, the other best is 1.03. At the same time, we also found that the relationship of dependency between different time series and which impact for short-term forecasting. Namely, the continues sequences on the short-term forecasting have good results while the discrete sequences consequence is bad. However, it can also strengthen the final performance on short-term load forecasting when we combine them by deep learning train process.
43
Huang, Bang-Xuan, and 黃邦烜. "Recurrent Neural Network-based Language Modeling with Extra Information Cues for Speech Recognition." Thesis, 2011. http://ndltd.ncl.edu.tw/handle/06701731515809147714.
Full textAbstract:
碩士<br>國立臺灣師範大學<br>資訊工程研究所<br>100<br>The goal of language modeling (LM) attempts to capture the regularities of natural languages. It uses large amounts of training text for model training so as to help predict the most likely upcoming word given a word history. Therefore, it plays an indispensable role in automatic speech recognition (ASR). The N-gram language model, which determines the probability of an upcoming word given its preceding N-1 word history, is most prominently used. When N is small, a typical N-gram language model lacks the ability of rendering long-span lexical information. On the other hand, when N becomes larger, it will suffer from the data sparseness problem because of insufficient training data. With this acknowledged, research on the neural network-based language model (NNLM), or more specifically, the feed-forward NNLM, has attracted considerable attention of researchers and practitioners in recent years. This is attributed to the fact that the feed-forward NNLM can mitigate the data sparseness problem when estimating the probability of an upcoming word given its corresponding word history through mapping them into a continuous space. In addition to the feed-forward NNLM, a recent trend is to use the recurrent neural network-based language model (RNNLM) to construct the language model for ASR, which can make efficient use of the long-span lexical information inherent in the word history in a recursive fashion.
In this thesis, we not only investigate to leverage extra information relevant to the word history for RNNLM, but also devise a dynamic model estimation method to obtain an utterance-specific RNNLM. We experimentally observe that our proposed methods can show promise and perform well when compared to the existing LM methods on a large vocabulary continuous speech recognition (LVCSR) task.
44
Lee, Chi-Wei, and 李啟為. "On-Line Portfolio Selection Using Recurrent Neural Network Based on Moving Average Reversion." Thesis, 2016. http://ndltd.ncl.edu.tw/handle/08650477242803367606.
Full textAbstract:
碩士<br>國立臺灣大學<br>資訊工程學研究所<br>104<br>As the development of computing power and the neural network methods become more and more mature, the prediction on financial instruments using artificial related methods becomes more and more viable. In this thesis, we combined the moving average reversion and recurrent neural network on online portfolio selection, which is one of the most popular topics in the area of financial industry. Moving average reversion technique has been proved powerful on online portfolio selection. On-line portfolio selection, in short, is to predict the stocks’prices of the next trading day and adjust the previous portfolio accordingly, while moving average reversion exploits the mean reversion characteristic of stock to better maximize the total wealth in the final trading day. Recurrent neural network (RNN) has been attracting increasing interest in time series prediction like handwriting recognition and speech recognition. RNNs use their internal memory to better utilize the historical sequences, which follows other time series prediction methods. Combining RNN and moving average reversion method, we’ve shown some improvement on popular datasets.
45
Li, Chun-Yi, and 李俊毅. "Design and Implementation of Recurrent-Neural-Network Based Controller for Maglev Suspension System." Thesis, 2016. http://ndltd.ncl.edu.tw/handle/20987867948465320138.
Full textAbstract:
碩士<br>中國文化大學<br>機械工程學系數位機電碩士班<br>104<br>There exist mechanical vibration, friction and wearing loss caused by contact operation in the conventional mechanical shock absorbers. Maglev suspension system (MSS), using the electromagnetic force to float, can effectively reduce above drawbacks. However the parameters in the mathematical model are related to the permanent magnet geometry, distance and the total mass of the platform. To achieve better response, a recurrent neural network (RNN) model control architecture for MSS is proposed to replace the conventional shock absorber in this thesis. Finally the proposed MSS is utilized to the six-foot robot.
With the extension of the applications of the MSS, the mathematical model of the system is full of uncertainties. Conventional controllers, currently used in the industry, cannot adapt to the complex environment, although their theory and architecture are simple. Several design methods based on the intelligent control have been announced; however, the researches of smart MSS are relatively rare. To deal with this problem, a RNN model is proposed as main controller and an auxiliary proportional-differential (PD) controller is added in the proposed architecture. By using the gathered data pairs, the more accurate model can be established in the changing environments.
To deal with the highly nonlinear dynamic system, the trained RNN is developed as the model of the MSS and the least-mean-square (LMS) error learning algorithm is proposed to tune the parameters. To further tackle the larger uncertainties, an auxiliary PD control effort is added. Thus, the fast response can be obtained without degrading the tracking performance. Some MATLAB simulated results, experimental results and one implemented MSS prototype are provided to verify the effectiveness of the proposed architecture.
46
Huang, Po-Kai, and 黃柏凱. "Recurrent Fuzzy Neural Network Controlled Linear Induction Motor Drive Based on Genetic Algorithm." Thesis, 2003. http://ndltd.ncl.edu.tw/handle/01832501290796089644.
Full textAbstract:
碩士<br>中原大學<br>電機工程研究所<br>91<br>The subject of this thesis is to develop a recurrent fuzzy neural network (RFNN) controlled linear induction motor (LIM) drive based on genetic algorithm (GA). First, the dynamic model of an indirect field-oriented linear induction motor drive is derived. The personal computer (PC) controlled LIM drive system consists of PC, AD/DA servo control card, and a ramp comparison current-controlled PWM. Then, the on-line tuning of the RFNN learning rates and weights using real-coded GA are developed individually to achieve the robust and precise position control of the LIM. Using real-coded GA to search the optimal learning rates of the RFNN can reduce time consumption in the trial-and-error process and guarantee an optimal solution. In addition, using the real-coded GA to search the optimal weights of the RFNN can avoid the network from converging slowly or diverging due to bad network architecture. Finally, the effectiveness of the proposed control schemes is demonstrated by simulated and experimental results.
47
CHEN, JUN-HONG, and 陳駿宏. "An Emergencies Predictable Time Series Prediction Model Based on Alternative Recurrent Neural Network." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/q8kby3.
Full textAbstract:
碩士<br>逢甲大學<br>資訊工程學系<br>106<br>Most of the algorithms commonly used for time series analysis and prediction are prone to unexpected failure when applied to real-world data, particularly when the inputs are beyond the range of the model, as occurs in many unexpected events. In this study, we sought to develop a model that is easy to train and applicable to a wide range of emergency situations. The proposed scheme uses a recurrent neural network model to detect potential emergencies through the analysis of time series data. In the event of an emergency, supplementary plug-in modules can be used to adjust the outputs. A back-propagation / genetic algorithm is used to train the model. Experiment results demonstrate the efficacy of the proposed model in the detection of emergency situations and dispatch of rapid response.
48
Lee, ChaoFa, and 李昭法. "A Position Controller Based on Diagonal Recurrent Neural Network for Shoulder Joint FES Applications." Thesis, 1999. http://ndltd.ncl.edu.tw/handle/96171475240367465197.
Full textAbstract:
碩士<br>國立臺灣大學<br>電機工程學研究所<br>87<br>This study proposed a novel diagonal recurrent neural network (DRNN) based controller for shoulder joint angle control in FES application. The controller, which contained a self recurrent feedback loops in each hidden node, could adopted its weighting from a modified error back-propagation algorithm. Joint angle, and the previous stimulation amplitude were fed to the controller to calculate present stimulation in real time.
Six able-bodied subjects were involved in the clinical trial. FES was applied over shoulder muscle groups (deltoid & supraspinatus muscles) to perform shoulder flexion and abduction motion. Different control schemes including traditional proportional controller, DRNN controller, and fix-adjustment assisted DRNN controller were adopted for comparison purposes. Performance was compared over joint angle tracking abilities of all schemes. The results showed that DRNN could dynamically learn the time-variant muscle properties during electrical stimulation (RMS error: 6.28 degrees), while traditional proportional controller's performance degraded while muscle adaptation occurred (RMS error: 9.32 degrees). Response delay did exist in DRNN controllers; therefore, a cooperative fix-adjustment controller can improve this drawback by offset stimulation current before the neural network converges (RMS error: 3.38 degrees),.
49
Wang, Te-Kai, and 王德楷. "Estimation on Return for Index: Based on Recurrent Neural Network and Modern Portfolio Theory." Thesis, 2017. http://ndltd.ncl.edu.tw/handle/gwr7dg.
Full textAbstract:
碩士<br>國立臺灣大學<br>資訊工程學研究所<br>105<br>Algorithms for trading strategy has been a hot topic for several years, and several algorithms have been introduced and claim to have a good performance. In our surveys, we found that most of the portfolio selection algorithms could be categorized into two categories. One of which relies on the financial background knowledge, and tries to implement the idea by program. The other way is to use existing method such as reinforcement learning or recurrent neural network, since RL (reinforcement learning) and RNNs (Recurrent Neural Network) are believed to be more effective in processing sequential data.
RNNs have already successfully caught people’s eyes in time series prediction, such as semantic analysis or translation. In our work, we tried to develop a strategy on stock market with RNN, suggest when is the good timing to purchase. However, the general model often suffers from bear market. Therefore, we tried to combine the Modern Portfolio Theory (MPT) into our work to lower the risk of our investment, and hope to make a higher profit.
50
Chen, Jia-Siang, and 陳家翔. "Constructing Embedding-based Recurrent Neural Network Recommendation System Using User Behavior and Product Information." Thesis, 2019. http://ndltd.ncl.edu.tw/handle/22am8x.
Full text