Um die anderen Arten von Veröffentlichungen zu diesem Thema anzuzeigen, folgen Sie diesem Link: Evolutionary development of neural network.
Dissertationen zum Thema „Evolutionary development of neural network“
Geben Sie eine Quelle nach APA, MLA, Chicago, Harvard und anderen Zitierweisen an
Machen Sie sich mit Top-50 Dissertationen für die Forschung zum Thema "Evolutionary development of neural network" bekannt.
Neben jedem Werk im Literaturverzeichnis ist die Option "Zur Bibliographie hinzufügen" verfügbar. Nutzen Sie sie, wird Ihre bibliographische Angabe des gewählten Werkes nach der nötigen Zitierweise (APA, MLA, Harvard, Chicago, Vancouver usw.) automatisch gestaltet.
Sie können auch den vollen Text der wissenschaftlichen Publikation im PDF-Format herunterladen und eine Online-Annotation der Arbeit lesen, wenn die relevanten Parameter in den Metadaten verfügbar sind.
Sehen Sie die Dissertationen für verschiedene Spezialgebieten durch und erstellen Sie Ihre Bibliographie auf korrekte Weise.
1
Bush, Brian O. „Development of a fuzzy system design strategy using evolutionary computation“. Ohio : Ohio University, 1996. http://www.ohiolink.edu/etd/view.cgi?ohiou1178656308.
Der volle Inhalt der Quelle
2
Townsend, Joseph Paul. „Artificial development of neural-symbolic networks“. Thesis, University of Exeter, 2014. http://hdl.handle.net/10871/15162.
Der volle Inhalt der QuelleAnnotation:
Artificial neural networks (ANNs) and logic programs have both been suggested as means of modelling human cognition. While ANNs are adaptable and relatively noise resistant, the information they represent is distributed across various neurons and is therefore difficult to interpret. On the contrary, symbolic systems such as logic programs are interpretable but less adaptable. Human cognition is performed in a network of biological neurons and yet is capable of representing symbols, and therefore an ideal model would combine the strengths of the two approaches. This is the goal of Neural-Symbolic Integration [4, 16, 21, 40], in which ANNs are used to produce interpretable, adaptable representations of logic programs and other symbolic models. One neural-symbolic model of reasoning is SHRUTI [89, 95], argued to exhibit biological plausibility in that it captures some aspects of real biological processes. SHRUTI's original developers also suggest that further biological plausibility can be ascribed to the fact that SHRUTI networks can be represented by a model of genetic development [96, 120]. The aims of this thesis are to support the claims of SHRUTI's developers by producing the first such genetic representation for SHRUTI networks and to explore biological plausibility further by investigating the evolvability of the proposed SHRUTI genome. The SHRUTI genome is developed and evolved using principles from Generative and Developmental Systems and Artificial Development [13, 105], in which genomes use indirect encoding to provide a set of instructions for the gradual development of the phenotype just as DNA does for biological organisms. This thesis presents genomes that develop SHRUTI representations of logical relations and episodic facts so that they are able to correctly answer questions on the knowledge they represent. The evolvability of the SHRUTI genomes is limited in that an evolutionary search was able to discover genomes for simple relational structures that did not include conjunction, but could not discover structures that enabled conjunctive relations or episodic facts to be learned. Experiments were performed to understand the SHRUTI fitness landscape and demonstrated that this landscape is unsuitable for navigation using an evolutionary search. Complex SHRUTI structures require that necessary substructures must be discovered in unison and not individually in order to yield a positive change in objective fitness that informs the evolutionary search of their discovery. The requirement for multiple substructures to be in place before fitness can be improved is probably owed to the localist representation of concepts and relations in SHRUTI. Therefore this thesis concludes by making a case for switching to more distributed representations as a possible means of improving evolvability in the future.
3
Hytychová, Tereza. „Evoluční návrh neuronových sítí využívající generativní kódování“. Master's thesis, Vysoké učení technické v Brně. Fakulta informačních technologií, 2021. http://www.nusl.cz/ntk/nusl-445478.
Der volle Inhalt der QuelleAnnotation:
The aim of this work is to design and implement a method for the evolutionary design of neural networks with generative encoding. The proposed method is based on J. F. Miller's approach and uses a brain model that is gradually developed and which allows extraction of traditional neural networks. The development of the brain is controlled by programs created using cartesian genetic programming. The project was implemented in Python with the use of Numpy library. Experiments have shown that the proposed method is able to construct neural networks that achieve over 90 % accuracy on smaller datasets. The method is also able to develop neural networks capable of solving multiple problems at once while slightly reducing accuracy.
4
Kadiyala, Akhil. „Development and Evaluation of an Integrated Approach to Study In-Bus Exposure Using Data Mining and Artificial Intelligence Methods“. University of Toledo / OhioLINK, 2012. http://rave.ohiolink.edu/etdc/view?acc_num=toledo1341257080.
Der volle Inhalt der Quelle
5
Adams, Bryan (Bryan Paul) 1977. „Evolutionary, developmental neural networks for robust robotic control“. Thesis, Massachusetts Institute of Technology, 2006. http://hdl.handle.net/1721.1/37900.
Der volle Inhalt der QuelleAnnotation:
Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2006.Includes bibliographical references (p. 136-143).
The use of artificial evolution to synthesize controllers for physical robots is still in its infancy. Most applications are on very simple robots in artificial environments, and even these examples struggle to span the "reality gap," a name given to the difference between the performance of a simulated robot and the performance of a.real robot using the same evolved controller. This dissertation describes three methods for improving the use of artificial evolution as a tool for generating controllers for physical robots. First, the evolutionary process must incorporate testing on the physical robot. Second, repeated structure on the robot should be exploited. Finally, prior knowledge about the robot and task should be meaningfully incorporated. The impact of these three methods, both in simulation and on physical robots, is demonstrated, quantified, and compared to hand-designed controllers.
by Bryan Adams.
Ph.D.
6
Tsui, Kwok Ching. „Neural network design using evolutionary computing“. Thesis, King's College London (University of London), 1998. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.299918.
Der volle Inhalt der Quelle
7
Hayward, Serge. „Financial forecasting and modelling with an evolutionary artificial neural network“. Thesis, Queen Mary, University of London, 2006. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.439394.
Der volle Inhalt der Quelle
8
Hlynka, Markian D. „A framework for an automated neural network designer using evolutionary algorithms“. Thesis, National Library of Canada = Bibliothèque nationale du Canada, 1999. http://www.collectionscanada.ca/obj/s4/f2/dsk1/tape7/PQDD_0014/MQ41716.pdf.
Der volle Inhalt der Quelle
9
Jagadeesan, Ananda Prasanna. „Real time evolutionary algorithms in robotic neural control systems“. Thesis, Robert Gordon University, 2006. http://hdl.handle.net/10059/436.
Der volle Inhalt der QuelleAnnotation:
This thesis describes the use of a Real-Time Evolutionary Algorithm (RTEA) to optimise an Artificial Neural Network (ANN) on-line (in this context “on-line” means while it is in use). Traditionally, Evolutionary Algorithms (Genetic Algorithms, Evolutionary Strategies and Evolutionary Programming) have been used to train networks before use - that is “off-line,” as have other learning systems like Back-Propagation and Simulated Annealing. However, this means that the network cannot react to new situations (which were not in its original training set). The system outlined here uses a Simulated Legged Robot as a test-bed and allows it to adapt to a changing Fitness function. An example of this in reality would be a robot walking from a solid surface onto an unknown surface (which might be, for example, rock or sand) while optimising its controlling network in real-time, to adjust its locomotive gait, accordingly. The project initially developed a Central Pattern Generator (CPG) for a Bipedal Robot and used this to explore the basic characteristics of RTEA. The system was then developed to operate on a Quadruped Robot and a test regime set up which provided thousands of real-environment like situations to test the RTEA’s ability to control the robot. The programming for the system was done using Borland C++ Builder and no commercial simulation software was used. Through this means, the Evolutionary Operators of the RTEA were examined and their real-time performance evaluated. The results demonstrate that a RTEA can be used successfully to optimise an ANN in real-time. They also show the importance of Neural Functionality and Network Topology in such systems and new models of both neurons and networks were developed as part of the project. Finally, recommendations for a working system are given and other applications reviewed.
10
Jakobsson, Henrik. „Inversion of an Artificial Neural Network Mapping by Evolutionary Algorithms with Sharing“. Thesis, University of Skövde, Department of Computer Science, 1998. http://urn.kb.se/resolve?urn=urn:nbn:se:his:diva-165.
Der volle Inhalt der QuelleAnnotation:
Inversion of the artificial neural network mapping is a relatively unexplored field of science. By inversion we mean that a search is conducted to find what input patterns that corresponds to a specific output pattern according to the analysed network. In this report, an evolutionary algorithm is proposed to conduct the search for input patterns. The hypothesis is that the inversion with the evolutionary search-method will result in multiple, separate and equivalent input patterns and not get stuck in local optima which possibly would cause the inversion to result in erroneous answer. Beside proving the hypothesis, the tests are also aimed at explaining the nature of inversion and how the result of inversion should be interpreted. At the end of the document a long list of proposed future work is suggested. Work, which might result in a deeper understanding of what the inversion means and maybe an automated analysis tool, based on inversion.
11
Chan, Heather Y. „Gene Network Inference and Expression Prediction Using Recurrent Neural Networks and Evolutionary Algorithms“. BYU ScholarsArchive, 2010. https://scholarsarchive.byu.edu/etd/2648.
Der volle Inhalt der QuelleAnnotation:
We demonstrate the success of recurrent neural networks in gene network inference and expression prediction using a hybrid of particle swarm optimization and differential evolution to overcome the classic obstacle of local minima in training recurrent neural networks. We also provide an improved validation framework for the evaluation of genetic network modeling systems that will result in better generalization and long-term prediction capability. Success in the modeling of gene regulation and prediction of gene expression will lead to more rapid discovery and development of therapeutic medicine, earlier diagnosis and treatment of adverse conditions, and vast advancements in life science research.
12
Smith, Duncan. „An evolutionary approach to optimising neural network predictors for passive sonar target tracking“. Thesis, Loughborough University, 2009. https://dspace.lboro.ac.uk/2134/26870.
Der volle Inhalt der QuelleAnnotation:
Object tracking is important in autonomous robotics, military applications, financial time-series forecasting, and mobile systems. In order to correctly track through clutter, algorithms which predict the next value in a time series are essential. The competence of standard machine learning techniques to create bearing prediction estimates was examined. The results show that the classification based algorithms produce more accurate estimates than the state-of-the-art statistical models. Artificial Neural Networks (ANNs) and K-Nearest Neighbour were used, demonstrating that this technique is not specific to a single classifier.
13
Fischer, Manfred M., und Yee Leung. „A Genetic Algorithm Based Evolutionary Computational Neural Network for Modelling Spatial Interaction Data“. WU Vienna University of Economics and Business, 1998. http://epub.wu.ac.at/4151/1/WSG_DP_6198.pdf.
Der volle Inhalt der QuelleAnnotation:
Building a feedforward computational neural network model (CNN) involves two distinct
tasks: determination of the network topology and weight estimation. The specification of a
problem adequate network topology is a key issue and the primary focus of this contribution.
Up to now, this issue has been either completely neglected in spatial application domains, or
tackled by search heuristics (see Fischer and Gopal 1994). With the view of modelling
interactions over geographic space, this paper considers this problem as a global
optimization problem and proposes a novel approach that embeds backpropagation learning
into the evolutionary paradigm of genetic algorithms. This is accomplished by interweaving a
genetic search for finding an optimal CNN topology with gradient-based backpropagation
learning for determining the network parameters. Thus, the model builder will be relieved of
the burden of identifying appropriate CNN-topologies that will allow a problem to be solved
with simple, but powerful learning mechanisms, such as backpropagation of gradient descent
errors. The approach has been applied to the family of three inputs, single hidden layer,
single output feedforward CNN models using interregional telecommunication traffic data for
Austria, to illustrate its performance and to evaluate its robustness. (authors' abstract)Series: Discussion Papers of the Institute for Economic Geography and GIScience
14
Westermann, Gert. „Constructivist neural network models of cognitive development“. Thesis, University of Edinburgh, 2000. http://hdl.handle.net/1842/22733.
Der volle Inhalt der QuelleAnnotation:
In this thesis I investigate the modelling of cognitive development with Constructivist neural networks. I argue that the constructivist nature of development, that is, the building of a cognitive system through active interactions with its environment, is an essential property of human development and should be considered in models of cognitive development. I evaluate this claim on the basis of evidence from cortical development, cognitive development, and learning theory. In an empirical evaluation of this claim, I then present a constructivist neural network model of the acquisition of the English past tense and of impaired inflectional processing in German agrammatic aphasics. The model displays a realistic course of acquisition, closely modelling the U-shaped learning curve and more detailed effects such as frequency and family effects. Further, the model develops double dissociations between regular and irregular verbs. I argue that the ability of the model to account for the human data is based on its constructivist nature, and this claim is backed by an analogous, but non-constructivist model that does not display many aspects of the human behaviour. Based on these results I develop a taxonomy for cognitive models that incorporates architectural and developmental aspects besides the traditional distinction between symbolic and subsymbolic processing. When the model is trained on the German participle and is then lesioned by removing connections, the breakdown in performance reflects the profiles of German aggrammatic aphasics. Irregular inflections are selectively impaired and are often overregularised. Further, frequency effects and the regularity-continuum effect that are observed in aphasic subjects can also be modelled. The model predicts that an aphasic profile with selectively impaired regular inflections would be evidence for a locally distinct processing of regular and irregular infections.
15
Yang, Yini. „Training Neural Networks with Evolutionary Algorithms for Flash Call Verification“. Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-283039.
Der volle Inhalt der QuelleAnnotation:
Evolutionary algorithms have achieved great performance among a wide range of optimization problems. In this degree project, the network optimization problem has been reformulated and solved in an evolved way. A feasible evolutionary framework has been designed and implemented to train neural networks in supervised learning scenarios. Under the structure of evolutionary algorithms, a well-defined fitness function is applied to evaluate network parameters, and a carefully derived form of approximate gradients is used for updating parameters. Performance of the framework has been tested by training two different types of networks, linear affine networks and convolutional networks, for a flash call verification task.Under this application scenario, whether a flash call verification will be successful or not will be predicted by a network, which is inherently a binary classification problem. Furthermore, its performance has also been compared with traditional backpropagation optimizers from two aspects: accuracy and time consuming. The results show that this framework is able to push a network training process to converge into a certain level. During the training process, despite of noises and fluctuations, both accuracies and losses converge roughly under the same pattern as in backpropagation. Besides, the evolutionary algorithm seems to have higher updating efficiency per epoch at the first training stage before converging. While with respect to fine tuning, it doesn’t work as good as backpropagation in the final convergence period.Evolutionära algoritmer uppnår bra prestanda för ett stort antal olika typer av optimeringsproblem. I detta examensprojekt har ett nätverksoptimeringsproblem lösts genom omformulering och vidareutveckling av angreppssättet. Ett förslag till ramverk har utformats och implementerats för att träna neuronnätverk i övervakade inlärningsscenarier. För evolutionära algoritmer används en väldefinierad träningsfunktion för att utvärdera nätverksparametrar, och en noggrant härledd form av approximerade gradienter används för att uppdatera parametrarna. Ramverkets prestanda har testats genom att träna två olika typer av linjära affina respektive konvolutionära neuronnätverk, för optimering av telefonnummerverifiering. I detta applikationsscenario förutses om en telefonnummerverifiering kommer att lyckas eller inte med hjälp av ett neuronnätverk som i sig är ett binärt klassificeringsproblem. Dessutom har dess prestanda också jämförts med traditionella backpropagationsoptimerare från två aspekter: noggrannhet och hastighet. Resultaten visar att detta ramverk kan driva en nätverksträningsprocess för att konvergera till en viss nivå. Trots brus och fluktuationer konvergerar både noggrannhet och förlust till ungefär under samma mönster som i backpropagation. Dessutom verkar den evolutionära algoritmen ha högre uppdateringseffektivitet per tidsenhet i det första träningsskedet innan den konvergerar. När det gäller finjustering fungerar det inte lika bra som backpropagation under den sista konvergensperioden.
16
Reiling, Anthony J. „Convolutional Neural Network Optimization Using Genetic Algorithms“. University of Dayton / OhioLINK, 2017. http://rave.ohiolink.edu/etdc/view?acc_num=dayton1512662981172387.
Der volle Inhalt der Quelle
17
Propst, Michael David. „Applying Linear Regression and Neural Network Meta-Models for Evolutionary Algorithm Based Simulation Optimization“. NCSU, 2009. http://www.lib.ncsu.edu/theses/available/etd-08112009-164218/.
Der volle Inhalt der QuelleAnnotation:
The increase in computing power over the last decade has led to an increase in the use of simulation programs to model real world optimization problems as well as the complexity with which these problems can be modeled. Once a model has been built, an experimental design is often used to determine the effects certain parameters have on the problem trying to determine the good settings that optimize a set of outputs. However, these problems often have a large number of variables or parameters that can be changed with wide value ranges and as these simulation models become increasingly more complex they become computationally expensive to run. Most of these problems are non-linear and may not have a true optimal solution based on the inherent variability in real-world applications and the stochastic simulation model. Evolutionary algorithms are a class of computational optimization techniques that harness the power of the computer to solve a problem. The application of evolutionary search techniques as a simulation optimization technique has yielded reasonable results. However, the algorithm can take a long time evaluating just one set of decision variables owing to replications and computational time of one simulation run and not to mention the sheer number of different sets that have to be evaluated to find good solutions for these complex problems. Linear regression and neural-network meta-models can be used to generate a surface model of the simulation. Evaluating the meta-model is very fast as compared to the simulation model. Therefore, this thesis combines the use of evolution algorithms, simulation models and meta-models to produce a more efficient simulation optimization technique. The two types of meta-models are tested to determine their effectiveness as a meta-modeling technique and the overall effectiveness of finding the best solution.
18
Sirois, Sylvain. „A neural network perspective on learning and development /“. Thesis, McGill University, 2000. http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=36836.
Der volle Inhalt der QuelleAnnotation:
This manuscript-based thesis explores the relationship between learning and development. The first manuscript reviews the important empirical regularities identified in human discrimination shift learning, including a qualitative age-related change in performance observed in childhood. Leading theoretical accounts of the empirical data are discussed, suggesting that none provides a comprehensive interpretation. The manuscript presents the novel, spontaneous overtraining interpretation. It hypothesizes that age-related changes in human shift learning stem from differences in amount of processing. Successful neural network simulations of the reversal and nonreversal shift tasks and of the optional shift task are reported as tests of the hypothesis.The second manuscript reports simulations of additional discrimination shift tasks. These are the intradimensional and extradimensional shift tasks, in which novel stimuli are introduced in the relearning phase. Preschoolers and adults exhibit the same pattern of behavior in this variant of shift learning. Simulation results show that the spontaneous overtraining hypothesis captures this effect.
The third chapter reports an empirical validation of the shift learning model. If the shift learning performance of adults is a consequence of more extensive processing, it follows that adults in whom such processing is prevented should perform as preschoolers. Sixty adults took part in a shift learning experiment with a Brown-Peterson task as a cognitive load. Results mirror those observed with preschoolers. As a control, 40 adults performed the shift learning experiment without the cognitive load. These results replicate the typical adult performance. Overall, these experiments lend additional support to the model developed in Manuscript 1.
The final manuscript is a theoretical discussion of the relationship between learning and development. Two classes of neural networks are discussed, and their underlying assumptions about learning and development are highlighted. These are static architecture and generative architecture networks. It is argued that only generative algorithms, such as used in the shift learning simulations, qualify as developmental models. Both classes of networks are further contrasted with respect to innateness. The comparison suggests that only generative networks can acquire genuinely new representations. The manuscript proposes a novel formulation of Piaget's constructivism from the generative neural network perspective.
19
Gessaroli, Erica <1983>. „Development, degeneration and neural network of the bodily self“. Doctoral thesis, Alma Mater Studiorum - Università di Bologna, 2014. http://amsdottorato.unibo.it/6342/.
Der volle Inhalt der QuelleAnnotation:
The question addressed by this dissertation is how the human brain builds a coherent representation of the body, and how this representation is used to recognize its own body. Recent approaches by neuroimaging and TMS revealed hints for a distinct brain representation of human body, as compared with other stimulus categories. Neuropsychological studies demonstrated that body-parts and self body-parts recognition are separate processes sub-served by two different, even if possibly overlapping, networks within the brain.
Bodily self-recognition is one aspect of our ability to distinguish between self and others and the self/other distinction is a crucial aspect of social behaviour. This is the reason why I have conducted a series of experiment on subjects with everyday difficulties in social and emotional behaviour, such as patients with autism spectrum disorders (ASD) and patients with Parkinson’s disease (PD). More specifically, I studied the implicit self body/face recognition (Chapter 6) and the influence of emotional body postures on bodily self-processing in TD children as well as in ASD children (Chapter 7). I found that the bodily self-recognition is present in TD and in ASD children and that emotional body postures modulate self and others’ body processing.
Subsequently, I compared implicit and explicit bodily self-recognition in a neuro-degenerative pathology, such as in PD patients, and I found a selective deficit in implicit but not in explicit self-recognition (Chapter 8). This finding suggests that implicit and explicit bodily self-recognition are separate processes subtended by different mechanisms that can be selectively impaired.
If the bodily self is crucial for self/other distinction, the space around the body (personal space) represents the space of interaction and communication with others. When, I studied this space in autism, I found that personal space regulation is impaired in ASD children (Chapter 9).
20
Fieldsend, Jonathan E. „Novel algorithms for multi-objective search and their application in multi-objective evolutionary neural network training“. Thesis, University of Exeter, 2003. http://hdl.handle.net/10871/11706.
Der volle Inhalt der Quelle
21
Van, Lierde Boris. „Developing Box-Pushing Behaviours Using Evolutionary Robotics“. Thesis, Högskolan Dalarna, Datateknik, 2011. http://urn.kb.se/resolve?urn=urn:nbn:se:du-6250.
Der volle Inhalt der QuelleAnnotation:
The context of this report and the IRIDIA laboratory are described in the preface. Evolutionary Robotics and the box-pushing task are presented in the introduction.The building of a test system supporting Evolutionary Robotics experiments is then detailed. This system is made of a robot simulator and a Genetic Algorithm. It is used to explore the possibility of evolving box-pushing behaviours. The bootstrapping problem is explained, and a novel approach for dealing with it is proposed, with results presented.Finally, ideas for extending this approach are presented in the conclusion.
22
Sung, Woong Je. „A neural network construction method for surrogate modeling of physics-based analysis“. Diss., Georgia Institute of Technology, 2012. http://hdl.handle.net/1853/43721.
Der volle Inhalt der QuelleAnnotation:
A connectivity adjusting learning algorithm, Optimal Brain Growth (OBG) was proposed. Contrast to the conventional training methods for the Artificial Neural Network (ANN) which focus on the weight-only optimization, the OBG method trains both weights and connectivity of a network in a single training process. The standard Back-Propagation (BP) algorithm was extended to exploit the error gradient information of the latent connection whose current weight has zero value. Based on this, the OBG algorithm makes a rational decision between a further adjustment of an existing connection weight and a creation of a new connection having zero weight. The training efficiency of a growing network is maintained by freezing stabilized connections in the further optimization process. A stabilized computational unit is also decomposed into two units and a particular set of decomposition rules guarantees a seamless local re-initialization of a training trajectory. The OBG method was tested for the multiple canonical, regression and classification problems and for a surrogate modeling of the pressure distribution on transonic airfoils. The OBG method showed an improved learning capability in computationally efficient manner compared to the conventional weight-only training using connectivity-fixed Multilayer Perceptrons (MLPs).
23
Cottens, Pablo Eduardo Pereira de Araujo. „Development of an artificial neural network architecture using programmable logic“. Universidade do Vale do Rio dos Sinos, 2016. http://www.repositorio.jesuita.org.br/handle/UNISINOS/5411.
Der volle Inhalt der QuelleAnnotation:
Submitted by Silvana Teresinha Dornelles Studzinski (sstudzinski) on 2016-06-29T14:42:16Z
No. of bitstreams: 1
Pablo Eduardo Pereira de Araujo Cottens_.pdf: 1315690 bytes, checksum: 78ac4ce471c2b51e826c7523a01711bd (MD5)Made available in DSpace on 2016-06-29T14:42:16Z (GMT). No. of bitstreams: 1 Pablo Eduardo Pereira de Araujo Cottens_.pdf: 1315690 bytes, checksum: 78ac4ce471c2b51e826c7523a01711bd (MD5) Previous issue date: 2016-03-07
Nenhuma
Normalmente Redes Neurais Artificiais (RNAs) necessitam estações de trabalho para o seu processamento, por causa da complexidade do sistema. Este tipo de arquitetura de processamento requer que instrumentos de campo estejam localizados na vizinhança da estação de trabalho, caso exista a necessidade de processamento em tempo real, ou que o dispositivo de campo possua como única tarefa a de coleta de dados para processamento futuro. Este projeto visa criar uma arquitetura em lógica programável para um neurônio genérico, no qual as RNAs podem fazer uso da natureza paralela de FPGAs para executar a aplicação de forma rápida. Este trabalho mostra que a utilização de lógica programável para a implementação de RNAs de baixa resolução de bits é viável e as redes neurais, devido à natureza paralelizável, se beneficiam pela implementação em hardware, podendo obter resultados de forma muito rápida.
Currently, modern Artificial Neural Networks (ANN), according to their complexity, require a workstation for processing all their input data. This type of processing architecture requires that the field device is located somewhere in the vicintity of a workstation, in case real-time processing is required, or that the field device at hand will have the sole task of collecting data for future processing, when field data is required. This project creates a generic neuron architecture in programmabl logic, where Artifical Neural Networks can use the parallel nature of FPGAs to execute applications in a fast manner, albeit not using the same resolution for its otputs. This work shows that the utilization of programmable logic for the implementation of low bit resolution ANNs is not only viable, but the neural network, due to its parallel nature, benefits greatly from the hardware implementation, giving fast and accurate results.
24
Lapedriza, Alberto. „Gene regulatory network of melanocyte development“. Thesis, University of Bath, 2016. https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.690724.
Der volle Inhalt der QuelleAnnotation:
Greenhill et al. (2011) developed a gene regulatory network of the main genes and interactions known to play a role in melanocyte biology, and generated a mathematical model to describe the behaviour of this complex network using semi quantitative data (ISH expression data). In this project we sought to collect expression data from four genes of the melanocyte GRN (sox10, kit, mitfa and dct) to develop a quantitative model that is able to describe the data more accurately. Moreover, we intended to identify more genes that are part of the melanocyte development process to be incorporated to the GRN. We analysed microarray data that compared differentially expressed genes between sox10 mutant and wild type embryos and validated five genes with a key role in melanocyte biology as downregulated in mitfa mutant embryos, which are downstream of mitfa in the GRN. We suggest that kit plays the role of factor Y in the Greenhill et al. (2011) GRN: Mitfa drives kit expression, and kit expression is transiently driven by Sox10 at early stages of development. As part of the feedback loop, kit seems to drive and maintain mitfa expression, however this needs to be validated. Finally we developed an experimental set up to obtain an estimate of gene expression per melanocyte from sox10, kit, mitfa and dct, using both qPCR and ISH cell count measurements. With this estimate we performed a parameter optimisation procedure, and found a set of parameters for the mathematical model that predicted the experimental data very accurately. The new model suggests that low expression values of sox10 are sufficient to drive mitfa expression in high levels. It also predicts that high expression of sox9b is needed to achieve the high expression levels of dct seen in the data, although these predictions need to be experimentally tested. This study represents the first attempt to obtain fine-scale gene expression data from melanocytes for the development of a quantitative mathematical model in zebrafish.
25
Southworth, David. „A simple artificial neural network development system for study and research“. Master's thesis, This resource online, 1991. http://scholar.lib.vt.edu/theses/available/etd-02162010-020120/.
Der volle Inhalt der Quelle
26
Bseiso, Abdallah. „DEVELOPMENT OF ARTIFICIAL NEURAL NETWORK SOFTWARE AND MODELS FOR ENGINEERING MATERIALS“. Cleveland State University / OhioLINK, 2021. http://rave.ohiolink.edu/etdc/view?acc_num=csu1610197805407219.
Der volle Inhalt der Quelle
27
Marshall, Stephen James. „The development of neural network based decision aids for clinical decision support“. Thesis, University of Sheffield, 1996. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.245570.
Der volle Inhalt der Quelle
28
Gualano, Leonardo. „Development of artificial neural network techniques for prediction of wheel-rail forces“. Thesis, Manchester Metropolitan University, 2007. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.436651.
Der volle Inhalt der QuelleAnnotation:
Over the past two decades, Artificial Neural Network (ANN) techniques have been used in many fields of research, due to their better robustness and fail ure tolerance capabilities compared with conventional modelling approaches. In applications such as railway vehicle dynamics, discrete scenarios of vehicle/track interactions are currently modelled using computer packages such as Simpack, ADAMS Rail, Vampire or Medyna, which use multi-body techniques to accurately model different aspects of rail vehicles and tracks such as derailment or passenger comfort. This thesis presents novel ANN techniques which make it possible to achieve ANN models accuracies comparable to those of multi-body techniques. A particular ANN structure has been designed with the aim of simplifying the training of ANNs with long training data sets. This is a Recurrent Neural Network (RNN) structure characterized by an optimized feedback technique which requires very little computational power. This novel structure and other more conventional RNN structures have been trained and tested and the processing times compared. The efficiency of the novel ANN structure in modelling a number of vehicle types, from passenger to friction damped freight vehicles, has been validated against commonly used techniques and also with a newly designed method, which consists of a combination of statistical functions applied to assess different aspects of the ANN models responses. The ANN designing and testing techniques have been implemented in a novel software tool, developed by the author, oriented to the investigation of ANN techniques for nonlinear system modelling. The novel ANN structure appears to be much faster than other structures commonly used for non-linear system modelling and adequate for the purposes of rail vehicle modelling. Compared to conventional ANN validation techniques, such as the mean square error and the cross-correlation function analysis, the novel assessment technique results in a more accurate quantification of the error terms and therefore, in a safer assessment which may be focused on aspects of the ANN model responses which are relevant in the context of railway engineering. The developed software tool is a significant contribution to the research by automating most of the repetitive tasks involved in the investigation and therefore allowing a considerable amount of ANNs to be designed and tested.
29
Kramer, Gregory Robert. „An analysis of neutral drift's effect on the evolution of a CTRNN locomotion controller with noisy fitness evaluation“. Wright State University / OhioLINK, 2007. http://rave.ohiolink.edu/etdc/view?acc_num=wright1182196651.
Der volle Inhalt der Quelle
30
Massera, Gianluca. „Evolution of grasping behaviour in anthropomorphic robotic arms with embodied neural controllers“. Thesis, University of Plymouth, 2012. http://hdl.handle.net/10026.1/1172.
Der volle Inhalt der QuelleAnnotation:
The works reported in this thesis focus upon synthesising neural controllers for anthropomorphic robots that are able to manipulate objects through an automatic design process based on artificial evolution. The use of Evolutionary Robotics makes it possible to reduce the characteristics and parameters specified by the designer to a minimum, and the robot’s skills evolve as it interacts with the environment. The primary objective of these experiments is to investigate whether neural controllers that are regulating the state of the motors on the basis of the current and previously experienced sensors (i.e. without relying on an inverse model) can enable the robots to solve such complex tasks. Another objective of these experiments is to investigate whether the Evolutionary Robotics approach can be successfully applied to scenarios that are significantly more complex than those to which it is typically applied (in terms of the complexity of the robot’s morphology, the size of the neural controller, and the complexity of the task). The obtained results indicate that skills such as reaching, grasping, and discriminating among objects can be accomplished without the need to learn precise inverse internal models of the arm/hand structure. This would also support the hypothesis that the human central nervous system (cns) does necessarily have internal models of the limbs (not excluding the fact that it might possess such models for other purposes), but can act by shifting the equilibrium points/cycles of the underlying musculoskeletal system. Consequently, the resulting controllers of such fundamental skills would be less complex. Thus, the learning of more complex behaviours will be easier to design because the underlying controller of the arm/hand structure is less complex. Moreover, the obtained results also show how evolved robots exploit sensory-motor coordination in order to accomplish their tasks.
31
Tepvorachai, Gorn. „An Evolutionary Platform for Retargetable Image and Signal Processing Applications“. Case Western Reserve University School of Graduate Studies / OhioLINK, 2008. http://rave.ohiolink.edu/etdc/view?acc_num=case1209504058.
Der volle Inhalt der Quelle
32
Kuchar, Olga Anna. „Development of animated finger movements via a neural network for tendon tension control“. Thesis, National Library of Canada = Bibliothèque nationale du Canada, 1999. http://www.collectionscanada.ca/obj/s4/f2/dsk3/ftp04/nq39322.pdf.
Der volle Inhalt der Quelle
33
Abou, Rayan Ihab Samy. „Development and evaluation of neural network models for cost reduction in unmanned air vehicles“. Thesis, University of Leicester, 2009. http://hdl.handle.net/2381/10880.
Der volle Inhalt der QuelleAnnotation:
With a growing demand for cost reduction in unmanned air vehicles (UAVs), there has been considerable interest in exploiting existing aircraft technologies. This thesis focuses on two technologies: model-based sensor fault detection, isolation and accommodation (SFDIA) schemes and flush air data sensing (FADS) systems. In the aerospace industry, SFDIA is traditionally based on physical (sensor) redundancy. Unfortunately this approach can be inadequate in UAVs due to cost, weight and space implications. Consequently researchers have found the concept of ‘virtual’ sensor redundancy (i.e. model-based methods) an invaluable alternative to physical redundancy. Current model-based SFDIA schemes rely on linear time-invariant (LTI) models. In nonlinear, time-varying systems (such as aircraft), LTI-based methods can sometimes fail to give satisfactory results. New approaches make use of neural network (NN) models due to their nonlinear and adaptive structure. In this thesis, a NN-based SFDIA scheme is designed to detect single and multiple sensor faults in a nonlinear UAV model. The proposed scheme has been shown to be robust to system and measurement noise and sensitive to a wide range of fault types. In the second part of this thesis, a FADS system is designed and tested on a mini air vehicle (MAV). With the primary goal of most air vehicle manufacturers being the reduction of costs, researchers found the concept of air data measurements using a matrix of pressure orifices to be a cheaper alternative to the standard air data boom. The concept of FADS systems is not new and has been quite popular in several NASA projects. However few applications consider MAVs where weight and cost implications can restrict the use of air data booms. The FADS system designed in this thesis has been shown to produce accurate air data estimations but more importantly has reduced instrumentation weight and cost by almost 80% and 97% respectively in comparison to a standard air data boom. The conclusions drawn from this thesis are clearly outlined at the end of each chapter and future work is also brought together in the final chapter.
34
Attal, Asadullah. „Development of Neural Network Models for Prediction of Highway Construction Cost and Project Duration“. Ohio University / OhioLINK, 2010. http://rave.ohiolink.edu/etdc/view?acc_num=ohiou1282146503.
Der volle Inhalt der Quelle
35
Schliebs, Stefan. „Heterogeneous probabilistic models for optimisation and modelling of evolving spiking neural networks“. AUT University, 2010. http://hdl.handle.net/10292/963.
Der volle Inhalt der QuelleAnnotation:
This thesis proposes a novel feature selection and classification method employing evolving spiking neural networks (eSNN) and evolutionary algorithms (EA). The method is named the Quantum-inspired Spiking Neural Network (QiSNN) framework. QiSNN represents an integrated wrapper approach. An evolutionary process evolves appropriate feature subsets for a given classification task and simultaneously optimises the neural and learning-related parameters of the network. Unlike other methods, the connection weights of this network are determined by a fast one-pass learning algorithm which dramatically reduces the training time. In its core, QiSNN employs the Thorpe neural model that allows the efficient simulation of even large networks. In QiSNN, the presence or absence of features is represented by a string of concatenated bits, while the parameters of the neural network are continuous. For the exploration of these two entirely different search spaces, a novel Estimation of Distribution Algorithm (EDA) is developed. The method maintains a population of probabilistic models specialised for the optimisation of either binary, continuous or heterogeneous search spaces while utilising a small and intuitive set of parameters. The EDA extends the Quantum-inspired Evolutionary Algorithm (QEA) proposed by Han and Kim (2002) and was named the Heterogeneous Hierarchical Model EDA (hHM-EDA). The algorithm is compared to numerous contemporary optimisation methods and studied in terms of convergence speed, solution quality and robustness in noisy search spaces. The thesis investigates the functioning and the characteristics of QiSNN using both synthetic feature selection benchmarks and a real-world case study on ecological modelling. By evolving suitable feature subsets, QiSNN significantly enhances the classification accuracy of eSNN. Compared to numerous other feature selection techniques, like the wrapper-based Multilayer Perceptron (MLP) and the Naive Bayesian Classifier (NBC), QiSNN demonstrates a competitive classification and feature selection performance while requiring comparatively low computational costs.
36
Svobodová, Jitka. „Neuronové sítě a evoluční algoritmy“. Master's thesis, Vysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií, 2009. http://www.nusl.cz/ntk/nusl-218221.
Der volle Inhalt der QuelleAnnotation:
Objective of this master's thesis is optimizing of neral network topology using some of evolutionary algorithms. The backpropagation neural network was optimized using genetic algorithms, evolutionary programming and evolutionary strategies. The text contains an application in the Matlab environment which applies these methods to simple tasks as pattern recognition and function prediction. Created graphs of fitness and error functions are included as a result of this thesis.
37
Schinazi, Robert Glen. „Designing Massive 3-Dimensional Neural Networks with Chromosomal-Based Simulated Development“. Diss., Virginia Tech, 1995. http://hdl.handle.net/10919/30531.
Der volle Inhalt der QuelleAnnotation:
A technique for designing and optimizing the next generation of smart process controllers has been developed in this dissertation. The literature review indicated that neural networks held the most promise for this application, yet fundamental limitations have prevented their introduction to commercial settings thus far. This fundamental limitation has been overcome through the enhancement of neural network theory.
The approach taken in this research was to produce highly intelligent process control systems by accurately modeling the nervous structures of higher biological organisms. The mammalian cerebral cortex was selected as the primary model since it is the only computational element capable of interpreting and complex patterns that develop over time. However the choice of the mammalian cerebral cortex as the model introduced two new levels of network complexity. First, the cerebral cortex is a three dimensional structure with extremely complicated patterns of interconnectivity. Second, the structure of the cerebral cortex can only be realized when thousands or millions of neurons are integrated into a massive scale neural network. The neural networks developed in this research were designed around the Hebbian adaptation, the only training technique proven by the literature review to be applicable to massive scale networks.
These design difficulties were resolved by not only modeling the cerebral cortex, but the process by which it develops and evolves in biological systems. To complete this model, an advanced genetic algorithm was produced, and a technique was developed to encode all functional and structural parameters that define the cerebral cortex into the artificial chromosome. The neural networks were designed by a cell growth simulation program that decoded the structural and functional information on the chromosome. The cell growth simulation program is capable of producing patterns of differentiation unique for any slight variations in the genetic parameters. These growth patterns are similar to patterns of cellular differentiation seen in biological systems. While the computational resources needed to implement a massive scale neural network are beyond that available in existing computer systems, the technique has produced output lists which fully define the interconnections and functional characteristic of the neurons, thereby laying the foundation for their future use in process control.Ph. D.
38
Kral, Zachary Tyler. „Development of an artificial neural network damage detection module for a structural health monitoring system“. Thesis, Wichita State University, 2009. http://hdl.handle.net/10057/2426.
Der volle Inhalt der QuelleAnnotation:
Aircraft, wind turbines, or space stations are expected to remain in service well beyond their designed
performance lifetime. Consequently, maintenance is an important issue for aircraft or aerospace structures. This is
accomplished through inspecting for damage at scheduled times and replacing damaged parts before failure. Ground
inspections of aircraft, even using simple nondestructive testing techniques, generally require the aircraft be pulled
from operation so that its components can be inspected for damage. Structural components are replaced if sufficient
damage is found. Research is underway to develop a structural health monitoring (SHM) system as a means to
improve the current maintenance routine. This system would consist of an array of sensors and associated analysis
codes which would scan for damage in-flight and perform real-time damage analysis of an aircraft’s structure. If
damage is recognized long before failure occurs, then a damage tolerance and prognostic assessment could be
implemented, allowing for a determination of the remaining life of components. The current method of inspecting
aircraft, consisting of ground inspections for damage after a set number of flight hours, works well from an aircraft
safety point of view. However, an in-flight SHM system would allow for better use of components, as specific
lifetimes could be determined; and, could be less costly, since an SHM system could be embedded into the aircraft
structure, thereby reducing or eliminating the need to tear down the aircraft to scan for damage during the ground
inspection and would ultimately lead to fewer required ground inspections.
General theory for material mechanics, fracture mechanics, waveform theory, and artificial neural networks are
presented in this paper. Among these, a simple triangulation method is developed to locate a crack tip position and a
procedure of combining the theory of fracture mechanics with waveform theories is introduced. These components
were used collectively in a series of experiments to investigate the possibility of using them in a future SHM system.
Flat aluminum panels, similar in thickness to those found in many aerospace structures, were subjected to
increasing static loading during laboratory tests. As the load increased, a designed crack in the panel increased in
size, releasing strain waves into the material. These waves were then detected by acoustic emission sensors, and
artificial neural networks were implemented to analyze the received strain waves. From a feed-forward neural
network, the crack length was approximated.
Next, similar aluminum panels were placed in a simply supported beam configuration with ultrasonic actuators
attached at various positions. These actuators created multiple point source locations, which was received by
multiple acoustic emission sensors. The location of the source was calculated by both triangulation method and an
vi
artificial neural network. A theory of plastic zone interference with strain waves released from crack tip extension
was introduced and shown as a possibility during the analysis of a final experiment. Sensors placed behind the crack
front were observed to detect waves with smaller amplitudes than the sensors placed in locations in front of the
crack during crack extension due to increasing, pseudo-static applied load. The effect of the acoustic emission
sensor placement relative to crack tip growth detections was determined to be possibly integrated with an artificial
neural network in future research.
Experiments were conducted to determine the crack length and location, using artificial neural network analyses
of acoustic emission signals. Artificial neural networks were developed which were trained with an existing dataset
of crack properties. These neural networks were applied to new situations that were not part of the training dataset.
The approximated crack growth of the artificial neural networks was around 10% shorter than the actual measured
crack growth length for an extension of around 0.8 in. Finally, some SHM analysis systems were proposed, based
on the conclusions made in the experiments. The artificial neural networks performed well at approximating both the
crack extension and location, using acoustic emission detections. These artificial neural networks in combination
with an acoustic emission system are reasonable candidate for the initial stages of a feasible structural health
monitoring system to determine the location and severity of structural damage within an aerospace structure.Thesis (M.S.)--Wichita State University, College of Engineering, Dept. of Aerospace Engineering
39
Thekkudan, Travis Francis. „Calibration of an Artificial Neural Network for Predicting Development in Montgomery County, Virginia: 1992-2001“. Thesis, Virginia Tech, 2008. http://hdl.handle.net/10919/33732.
Der volle Inhalt der QuelleAnnotation:
This study evaluates the effectiveness of an artificial neural network (ANN) to predict locations of urban change at a countywide level by testing various calibrations of the Land Transformation Model (LTM). It utilizes the Stuttgart Neural Network Simulator (SNNS), a common medium through which ANNs run a back-propagation algorithm, to execute neural net training.
This research explores the dynamics of socioeconomic and biophysical variables (derived from the 1990 Comprehensive Plan) and how they affect model calibration for Montgomery County, Virginia. Using NLCD Retrofit Land Use data for 1992 and 2001 as base layers for urban change, we assess the sensitivity of the model with policy-influenced variables from data layers representing road accessibility, proximity to urban lands, distance from urban expansion areas, slopes, and soils. Aerial imagery from 1991 and 2002 was used to visually assess changes at site-specific locations.
Results show a percent correct metric (PCM) of 32.843% and a Kappa value of 0.319. A relative operating characteristic (ROC) value of 0.660 showed that the model predicted locations of change better than chance (0.50). It performs consistently when compared to PCMs from a logistic regression model, 31.752%, and LTMs run in the absence of each driving variable ranging 27.971% â 33.494%. These figures are similar to results from other land use and land cover change (LUCC) studies sharing comparable landscape characteristics. Prediction maps resulting from LTM forecasts driven by the six variables tested provide a satisfactory means for forecasting change inside of dense urban areas and urban fringes for countywide urban planning.Master of Science
40
Edossa, D. C., und M. S. Babel. „Development of streamflow forecasting model using artificial neural network in the Awash River Basin, Ethiopia“. Interim : Interdisciplinary Journal, Vol 10 , Issue 1: Central University of Technology Free State Bloemfontein, 2011. http://hdl.handle.net/11462/332.
Der volle Inhalt der QuelleAnnotation:
Published ArticleEarly indication of possible drought can help in developing suitable drought mitigation strategies and measures in advance. Therefore, drought forecasting plays an important role in the planning and management of water resource in such circumstances. In this study, a non-linear streamflow forecasting model was developed using Artificial Neural Network (ANN) modeling technique at the Melka Sedi stream gauging station, Ethiopia, with adequate lead times. The available data was divided into two independent sets using a split sampling tool of the neural network software. The first data set was used for training and the second data set, which is normally about one fourth of the total available data, was used for testing the model. A one year data was set aside for validating the ANN model. The streamflow predicted using the model on weekly time step compared favorably with the measured streamflow data (R2 = 75%) during the validation period. Application of the model in assessing appropriate agricultural water management strategies for a large-scale irrigation scheme in the Awash River Basin, Ethiopia, has already been considered for publication in a referred journal.
41
Wu, Dong-Perng, und 吳東朋. „Evolutionary Neural Network Development and Optimization“. Thesis, 1996. http://ndltd.ncl.edu.tw/handle/90531512107411022471.
Der volle Inhalt der QuelleAnnotation:
碩士國立交通大學
資訊科學學系
84
Neural Networks and Genetic Algorithms are two kind of famous Computational Intelligence Models. They have shown their ability in many research domains. In this thesis, we combine these two technologies to solve pattern recognition problems. In biology, bacteria can change their expression of genes in order to save energy. Based on this observation, we combine a new approach called conditional genes with genetic algorithmsto achieve neural network structure self-organization. We applythis method to solve numerals recognition. Kohonen's Self-Organizing Feature Maps can adapt the output node topologically. We take advantages of this characteristic to adapt the structure of neural networks dynamically and a newrecognizing method is developed. After combining with conditional genes, neural networks can successfully recognize hand-written numerals.
42
Li, Keh-Tsong, und 李克聰. „Neural Network combined with Genetic Algorithm-Evolutionary Neural Network“. Thesis, 1999. http://ndltd.ncl.edu.tw/handle/37236508646662658444.
Der volle Inhalt der QuelleAnnotation:
碩士國立交通大學
電機與控制工程系
87
This thesis presents a Real-Coded Rank-Based Genetic Algorithm (RCRBGA), which is represented by a chromosome containing parameters in floating-point. The use of rank-based fitness increases the population diversity. The offspring are generated by the rank-based reproduction, real parametric crossover and mutation in the evolving process. Besides, an Evolutionary Neural Network (ENN) which combines RCRBGA and Back-Propagation (BP) is introduced. ENN applies the learning concept to the evolution process, like the behavior of human beings. It not only improves the disadvantage of easily slumping in to local minima of BP but also overcomes the defect of genetic algorithm, which can't efficiently converge to minima. Finally, the search ability of RCRBGA is demonstrated by an example, linear state-feedback controller via pole-assignment method. In addition, ENN applies to a classifying problem of the modified XOR to show its advantage.
43
阮呂正璽. „MeNet : A Multi-Objective Evolutionary Artificial Neural Network“. Thesis, 2002. http://ndltd.ncl.edu.tw/handle/47873963540026950168.
Der volle Inhalt der QuelleAnnotation:
碩士長庚大學
資訊管理研究所
90
Interest in algorithms which dynamically construct artificial neural networks has been growing in recent years. The traditional methods that have been used to construct near optimal artificial neural networks are to minimize the sequence of error functions associated with the growing network. But, constructing a near optimal artificial neural networks by considering the sequence of error unilaterally, will lead to more complicated network architecture. This paper proposes an evolutionary system for constructing neural networks named “MeNet”. Combining the concepts of “Evolutionary Programming” and ”Multi-Objective Programming”, MeNet minimizes the sequence of error and the complexity of network architecture simultaneously. MeNet has been tested on two benchmark problems, including Xor problem and 8-bit parity checking problem. The results show that MeNet can produce neural network architecture with lower complexity while keeping the acceptable accuracy. Such effects are clear, especially in testing more complicated problem.
44
李宜勳. „Evolutionary Learning and Application of BMF Fuzzy-Neural Network“. Thesis, 2002. http://ndltd.ncl.edu.tw/handle/13551677512428832309.
Der volle Inhalt der QuelleAnnotation:
碩士輔仁大學
電子工程學系
90
In this thesis, a novel approach to adjust both the control points of B-spline membership functions (BMFs) and the weightings of fuzzy-neural networks using a simplified genetic algorithm (SGA) is proposed. Fuzzy-neural networks (FNN) are traditionally trained by using gradient-based methods, and may fall into a local minimum during the learning process. Genetic algorithms have drawn significant attentions in various fields due to their capabilities of directed random search for global optimization. This motivates the use of genetic algorithms to overcome the problem encountered by the conventional learning methods. However, it is well known that searching speed of the conventional genetic algorithms is not desirable. Thus far, such conventional genetic algorithms are inherently disadvantaged in dealing with a vast amount (over 100) of adjustable parameters in the fuzzy-neural networks. In this thesis, the SGA is proposed by using a sequential-search-based crossover point (SSCP) method in which a better crossover point is determined and only the gene at the specified crossover point is crossed as a single point crossover operation. Chromosomes consisting of both the control points of BMF’s and the weightings of fuzzy-neural networks are coded as an adjustable vector with real number components and searched by the SGA. Adaptive control is a technique of applying some system identification techniques to obtain a model of a system from input-output data. GA on-line training is one of most significant studies of the subject. However, with traditional GA no further training of the FNN is possible, while the drive is operating since the training process is too slow for on-line training. Because of the use of the SGA, faster convergence of the evolution process to search for an optimal fuzzy-neural network can be achieved. Unknown systems identified by using the fuzzy-neural networks via the SGA are applied to adaptive fuzzy-neural control at the end of this thesis to illustrate the effectiveness and applicability of the proposed method.
45
Wang, Chao-Chi, und 王炤棋. „Evolutionary Computation and Parallel Neural Network for Channel Assignment“. Thesis, 1997. http://ndltd.ncl.edu.tw/handle/69306607788362415125.
Der volle Inhalt der Quelle
46
Tsai, Kuen-Yan, und 蔡昆洋. „An Evolutionary Approach toward Self-Organization of Neural Network“. Thesis, 1997. http://ndltd.ncl.edu.tw/handle/67772871322786701226.
Der volle Inhalt der QuelleAnnotation:
碩士國立交通大學
資訊科學學系
86
The construction of a proper network architecture is always a tough challenge while we apply neural networks. In this thesis, we propose a construction method inspired by biological nervous systems, which is developed based on the construction rules and information encoded in genes. According to this idea, we design the construction rules for neural networks and encode the rules. Then, evolve the optimal rules by using genetic algorithms.The spirit of rule construction is mainly from the competition rules for resource in the nature. Based on this hypothesis, neurons in our model need resources to survive, and input data are considered as the resources. After investigating the interaction between the neurons and the resources, we organize the general construction rules for neural networks.In this thesis, we prove the usefulness of the proposed model by constructing two different neural network models, RCE (Reduced Coulomb Energy) and Kohonen*s SOM(Self-Organizing Map),with the application of IRIS training data, and hand-written digit recognition. Simulation results show that the neural network architectures can be efficiently constructed by the proposed model.
47
楊世宏. „Neural Network with Evolutionary Structure Learning and Its Prediction Application“. Thesis, 2011. http://ndltd.ncl.edu.tw/handle/34255734360192730143.
Der volle Inhalt der QuelleAnnotation:
博士國立交通大學
電控工程研究所
100
This dissertation proposes a feedforward-neural-network-aided grey model (FNAGM) and its related on-line parameter learning and structure learning algorithms. The FNAGM uses a first-order single variable grey model (GM(1,1)) to predict signal and adopts a feedforward neural network (NN) to compensate the prediction error of GM(1,1). Furthermore, an on-line batch training is proposed to update the weights of NN in real-time. Thus, FNAGM can precisely predict and adapt itself to the dynamical change of the signal. To design the structure of FNAGM efficiently, a neuron-based structure learning, called symbiotic structure learning algorithm (SSLA), is proposed to establish the topology of NN. The SSLA constructs a neuron population and then builds a network population from the neuron population, and it can arbitrarily develop cascade NNs and feedforward NNs in an easy way. Further, SSLA carries out neuron crossover and mutation on the neuron population according to the idea of symbiotic evolution. The evolved FNAGM is applied to predict the signal and continuously adapt itself to the environment by the on-line batch training. On the other hand, a network-based structure learning, called evolutionary constructive and pruning algorithm (ECPA), is proposed to design the topology of NN by incorporating constructive and pruning methods in an evolutionary way. The ECPA starts from a set of NNs with the simplest possible structures, one hidden neuron connected to an input node. It then adds hidden neurons and connections by using the network crossover and mutation to increase the processing capabilities of NNs. Furthermore, a cluster-based pruning is proposed to prune insignificant neurons in a stochastic way. An age-based survival selection is proposed to delete old NNs with potentially complex structures and then introduce new NNs with the simplest possible structures. Numerical and experimental results of prediction problems show the effectiveness and feasibility of the proposed methods.
48
Wei-Chong, Cheng, und 鄭維崇. „The Study of Evolutionary Artificial Neural Network for Mobile Robot Navigation“. Thesis, 1998. http://ndltd.ncl.edu.tw/handle/51595137777333455129.
Der volle Inhalt der QuelleAnnotation:
碩士大同工學院
機械工程研究所
86
Different kind of intelligent learning rules have been applied todevelop intelligent robots recently. An intelligent robot should beable to learn to behave itself in an unknown environment using theinput information obtained from sensors. Clearly, the more complex theenvironment, the more difficult the learning task. Evolutionaryartificial neural network (EANN), a combination of artificial neuralnetwork and evolutionary search procedures, is applied in this study todevelop intelligent mobile robot due to it''s adaptability to changingenvironment. EANN is applied in this thesis to develop an intelligent mobilerobot which has the capability of learning navigation and obstacleavoidance in an unknown environment. A novel encoding, combining theconcepts of biological network and artificial neural network, isproposed to develop the EANN. With the aid of genetic algorithms, thecoded neural network''s architecture is able to learn itself. Tovalidate the performance of the proposed EANN, a GUI (graphic userinterface) mobile robot simulator written in Borland C++ is developed.The simulation results have shown the feasibility of the proposed EANN.
49
Sudjono, Erick, und 謝德祥. „Evolutionary Fuzzy Hybrid Neural Network for Decision-Making in Construction Management“. Thesis, 2008. http://ndltd.ncl.edu.tw/handle/75977105749001014871.
Der volle Inhalt der QuelleAnnotation:
碩士國立臺灣科技大學
營建工程系
96
Many studies have found that High-order neural network (HONN) is enables to boost neural network performance. This research utilize a hybrid model with HONN and Linear Neural Network (NN) concepts to develop high-order and linear neural connectors for layer connections. Consequently, this developed HNN will involve a linear/nonlinear switch for each neural layer connection. Furthermore, fuzzy logic (FL) has already been introduced to neural network and it is also found that the combination of FL and FNN has been proof-reading. Furthermore, fuzzy logic (FL) also has been introduced to neural network and been proofed with fuzzy neural network (FNN). Therefore, this research fuses fuzzy logic additionally to develop a fuzzy hybrid neural network (FHNN) architecture. Sequentially, genetic algorithm (GA) is employed to globally optimize membership function of FL and HNN topology and parameters. The fundamental of this developed model is a FHNN together with genetic optimization to develop the proposed evolutionary fuzzy hybrid neural networks (EFHNN). EFHNN is capable of handling complexity such as fuzzy/uncertain tasks, linear/nonlinear neural mapping and global optimization. Furthermore, in order to enable to process the EFHNN automatic adaptation, this research work integrates the EFHNN with object-oriented (OO) computer technique which consist of three modules: management module, adaptation module and inference module. The developed system was named as evolutionary fuzzy hybrid neural inference system (EFHNIS). The main focus of this research will be the optimum linear/nonlinear combinations for neural layers, which is the basis of the proposed hybrid neural network as well as the validation of the proposed EFHNN which collaborated with NN, HONN, FL, and GA concepts. In addition, since the construction managements might have certain issues which are complex and full of uncertain, implementing EFHNN in this topic is proved to be suitable and applicable to reliably assist the decision making in construction industry.
50
Magnier, Laurent. „Multiobjective optimization of building design using artificial neural network and multiobjective evolutionary algorithms“. Thesis, 2008. http://spectrum.library.concordia.ca/976218/1/MR63229.pdf.
Der volle Inhalt der QuelleAnnotation:
Building design is a very complex task, involving many parameters and conflicting objectives. In order to maximise the comfort and minimize the environmental impact, multiobjective optimization should be used. While some tools such as Genetic Algorithms (GA) exist, they are very seldom used in the industry, due to the large computational time they require. This thesis focuses on a specific approach called GAINN (Genetic Algorithm Integrating Neural Network), which combines the rapidity of evaluation of Artificial Neural Networks (ANN) with the optimization power of GAs. The thesis concentrates on a better handling of multiple objectives, in order to efficiently exploit the methodology and increase its accessibility for the non-expert. First, a Multiobjective Evolutionary Algorithm (MOEA), NSGA-II, has been selected and programmed in MATLAB. Then, two new MOEAs were developed, specifically designed to take advantage of GAINN fast evaluations. These two MOEAs have proven to be more efficient than NSGA-II on benchmark test functions, for a comparison based on a maximum runtime. In a second part of this thesis, developed MOEAs were used inside GAINN methodology to optimize the energy consumption and the thermal comfort in a residential building. This optimization was successful, and enabled significant improvements in terms of energy consumption and thermal comfort. It also enabled to illustrate very clearly the relation between these two objectives. This optimization however highlighted two limitations regarding the ANN, the number of training cases and the accuracy in the vicinity of optimal solutions. Finally, the developed algorithms were applied on a past optimization study, in order to highlight the improvements added to GAINN methodology by the use of MOEA.