Dissertations / Theses on the topic 'Perceptron'
To see the other types of publications on this topic, follow the link: Perceptron.
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 'Perceptron.'
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
Vieira, Douglas Alexandre Gomes. "Rede perceptron com camadas paralelas (PLP - Parallel Layer Perceptron)." Universidade Federal de Minas Gerais, 2006. http://hdl.handle.net/1843/BUOS-8CTH6W.
Full textAbstract:
This work presents a novel approach to deal with the structural risk minimization (SRM) applied to a general machine learning problem. The formulation is based on the fundamental concept that supervised learning is a bi-objective optimization problem in which two conflicting objectives should be minimized. The objectives are related to the training error, empirical risk (Remp), and the machine complexity (?). In this work one general Q-norm like method to compute the machine complexity is presented and it can be used to model and compare most of the learning machines found in the literature. The main advantage of the proposed complexity measure is that it is a simple method to split the linear and non-linear complexity influences, leading to a better understanding of the learning process. One novel learning machine, the Parallel Layer Perceptron (PLP) network was proposed here using a training algorithm based on the definitions and structures of learning, the Minimum Gradient Method (MGM). The combination of the PLP with the MGM (PLP-MGM) is held using a reliable least-squares procedure and it is the main contribution of this work.Este trabalho apresenta uma nova abordagem para lidar com o problema de minimização do risco estrutural (structural risk minimization - SRM) aplicado ao problema geral de aprendizado de máquinas. A formulação é baseada no conceito fundamental de que o aprendizado supervisionado é um problema de otimização bi-objetivo, onde dois objetivos conflitantes devem ser minimizados. Estes objetivos estão relacionados ao erro de treinamento, risco empírico (Remp), e à complexidade (capacidade) da máquina de aprendizado (?). Neste trabalho uma formulação geral baseada na norma-Q é utilizada para calcular a complexidade da máquina e esta pode ser utilizada para modelar e comparar a maioria das máquinas de aprendizado encontradas na literatura. A principal vantagem da medida proposta é que esta é uma maneira simples de separar as influências dos parâmetros lineares e não-lineares na medida de complexidade, levando a um melhor entendimento do processo de aprendizagem. Uma nova máquina de aprendizado, a rede perceptron com camadas paralelas (Parallel Layer Perceptron -PLP), foi proposta neste trabalho utilizando um treinamento baseado nas definições e estruturas de aprendizado propostas nesta tese, o Método do Gradiente Mínimo (Minimum Gradient Method-MGM). A combinação da PLP com o MGM (PLP-MGM) é feita utilizando o estimador de mínimos quadrados, sendo esta a principal contribuição deste trabalho.
2
Tsampouka, Petroula. "Perceptron-like large margin classifiers." Thesis, University of Southampton, 2007. https://eprints.soton.ac.uk/264242/.
Full textAbstract:
We address the problem of binary linear classification with emphasis on algorithms that lead to separation of the data with large margins. We motivate large margin classification from statistical learning theory and review two broad categories of large margin classifiers, namely Support Vector Machines which operate in a batch setting and Perceptron-like algorithms which operate in an incremental setting and are driven by their mistakes. We subsequently examine in detail the class of Perceptron-like large margin classifiers. The algorithms belonging to this category are further classified on the basis of criteria such as the type of the misclassification condition or the behaviour of the effective learning rate, i.e. the ratio of the learning rate to the length of the weight vector, as a function of the number of mistakes. Moreover, their convergence is examined with a prominent role in such an investigation played by the notion of stepwise convergence which offers the possibility of a rather unified approach. Whenever possible, mistake bounds implying convergence in a finite number of steps are derived and discussed. Two novel families of approximate maximum margin algorithms called CRAMMA and MICRA are introduced and analysed theoretically. In addition, in order to deal with linearly inseparable data a soft margin approach for Perceptron-like large margin classifiers is discussed. Finally, a series of experiments on artificial as well as real-world data employing the newly introduced algorithms are conducted allowing a detailed comparative assessment of their performance with respect to other well-known Perceptron-like large margin classifiers and state-of-the-art Support Vector Machines.
3
Shadafan, Raed Salem. "Sequential training of multilayer perceptron classifiers." Thesis, University of Cambridge, 1995. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.387686.
Full text
4
Dunne, R. A. "Multi-layer perceptron models for classification." Thesis, Dunne, R.A. (2003) Multi-layer perceptron models for classification. PhD thesis, Murdoch University, 2003. https://researchrepository.murdoch.edu.au/id/eprint/50257/.
Full textAbstract:
This thesis concerns the Multi-layer Perceptron (MLP) model, one of a variety of neural network models that have come into wide prominence since the mid 1980s for the classification of individuals into pre-defined classes based on a vector of individual measurements.
Each discipline in which the MLP model has had influence, including computing, electrical engineering and psychology, has recast the model into its own language and imbued it with its own concerns. This divergence of terminologies has made the literature somewhat impenetrable but has also led to an appreciation of other disciplines' priorities and interests.
The major aim of the thesis has been to bring the MLP model within the framework of statistics. We have two aims here: one is to make the MLP model more intelligible to statisticians; and the other is to bring the insights of statistics to the MLP model. A statistical modeling approach can make valuable contributions, ranging from small but important clarifications, such as clearing up the confusion in the MLP literature between the model and the methodology for fitting the model, to much larger insights such as determining the robustness of the model in the event of outlying or atypical data.
We provide a treatment of the relationship of the MLP classifier to more familiar statistical models and of the various fitting and model selection methodologies currently used for MLP models. A description of the influence curves of the MLP is provided, leading to both an understanding of how the MLP model relates to logistic regression (and to robust versions of logistic regression) and to a proposal for a robust MLP model.
Practical problems associated with the fitting of MLP models, from the effects of scaling of the input data to the effects of various penalty terms, are also considered. The MLP model has a variable architecture with the major source of variation being the number of hidden layer processing units. A direct method is given for determining this in multi-class problems where the pairwise decision boundary is linear in the feature space.
Finally, in applications such as remote sensing each vector of measurements or pixel contains contextual information about the neighboring pixels. The MLP model is modified to incorporate this contextual information into the classification procedure.
5
Rouleau, Christian. "Perceptron sous forme duale tronquée et variantes." Thesis, Université Laval, 2007. http://www.theses.ulaval.ca/2007/24492/24492.pdf.
Full textAbstract:
L’apprentissage automatique fait parti d’une branche de l’intelligence artificielle
et est utilisé dans de nombreux domaines en science. Il se divise en trois catégories
principales : supervisé, non-supervisé et par renforcement. Ce mémoire de maîtrise
portera uniquement sur l’apprentissage supervisé et plus précisément sur la classification
de données.
Un des premiers algorithmes en classification, le perceptron, fut proposé dans les
années soixante. Nous proposons une variante de cet algorithme, que nous appelons
le perceptron dual tronqué, qui permet l’arrêt de l’algorithme selon un nouveau critère.
Nous comparerons cette nouvelle variante à d’autres variantes du perceptron. De
plus, nous utiliserons le perceptron dual tronqué pour construire des classificateurs plus
complexes comme les «Bayes Point Machines».Machine Learning is a part of the artificial intelligence and is used in many fields in science. It is divided into three categories : supervised, not supervised and by reinforcement. This master’s paper will relate only the supervised learning and more precisely the classification of datas. One of the first algorithms in classification, the perceptron, was proposed in the Sixties. We propose an alternative of this algorithm, which we call the truncated dual perceptron, which allows the stop of the algorithm according to a new criterion. We will compare this new alternative with other alternatives of the perceptron. Moreover, we will use the truncated dual perceptron to build more complex classifiers like the «Bayes Point Machines».
6
Power, Phillip David. "Non-linear multi-layer perceptron channel equalisation." Thesis, Queen's University Belfast, 2001. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.343086.
Full text
7
Auld, Thomas James. "Bayesian applications of multilayer perceptron neural networks." Thesis, University of Cambridge, 2007. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.613209.
Full text
8
Kelby, Robin J. "Formalized Generalization Bounds for Perceptron-Like Algorithms." Ohio University / OhioLINK, 2020. http://rave.ohiolink.edu/etdc/view?acc_num=ohiou1594805966855804.
Full text
9
Midhall, Ruben, and Amir Parmbäck. "Utvärdering av Multilayer Perceptron modeller för underlagsdetektering." Thesis, Malmö universitet, Fakulteten för teknik och samhälle (TS), 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:mau:diva-43469.
Full textAbstract:
Antalet enheter som är uppkopplade till internet, Internet of Things (IoT), ökar hela tiden. År 2035 beräknas det finnas 1000 miljarder Internet of Things-enheter. Samtidigt som antalet enheter ökar, ökar belastningen på internet-nätverken som enheterna är uppkopplade till. Internet of Things-enheterna som finns i vår omgivning samlar in data som beskriver den fysiska tillvaron och skickas till molnet för beräkning. För att hantera belastningen på internet-nätverket flyttas beräkningarna på datan till IoT-enheten, istället för att skicka datan till molnet. Detta kallas för edge computing. IoT-enheter är ofta resurssnåla enheter med begränsad beräkningskapacitet. Detta innebär att när man designar exempelvis "machine learning"-modeller som ska köras med edge computing måste algoritmerna anpassas utifrån de resurser som finns tillgängliga på enheten. I det här arbetet har vi utvärderat olika multilayer perceptron-modeller för mikrokontrollers utifrån en rad olika experiment. "Machine learning"-modellerna har varit designade att detektera vägunderlag. Målet har varit att identifiera hur olika parametrar påverkar "machine learning"-systemen. Vi har försökt att maximera prestandan och minimera den mängd fysiskt minne som krävs av modellerna. Vi har även behövt förhålla oss till att mikrokontrollern inte haft tillgång till internet. Modellerna har varit ämnade att köras på en mikrokontroller "on the edge". Datainsamlingen skedde med hjälp av en accelerometer integrerad i en mikrokontroller som monterades på en cykel. I studien utvärderas två olika "machine learning"-system, ett som är en kombination av binära klassificerings modeller och ett multiklass klassificerings system som framtogs i ett tidigare arbete. Huvudfokus i arbetet har varit att träna modeller för klassificering av vägunderlag och sedan utvärdera modellerna. Datainsamlingen gjordes med en mikrokontroller utrustad med en accelerometer monterad på en cykel. Ett av systemen lyckas uppnå en träffsäkerhet på 93,1\% för klassificering av 3 vägunderlag. Arbetet undersöker även hur mycket fysiskt minne som krävs av de olika "machine learning"-systemen. Systemen krävde mellan 1,78kB och 5,71kB i fysiskt minne.The number of devices connected to the internet, the Internet of Things (IoT), is constantly increasing. By 2035, it is estimated to be 1,000 billion Internet of Things devices in the world. At the same time as the number of devices increase, the load on the internet networks to which the devices are connected, increases. The Internet of Things devices in our environment collect data that describes our physical environment and is sent to the cloud for computation. To reduce the load on the internet networks, the calculations are done on the IoT devices themselves instead of in the cloud. This way no data needs to be sent over the internet and is called edge computing. In edge computing, however, other challenges arise. IoT devices are often resource-efficient devices with limited computing capacity. This means that when designing, for example, machine learning models that are to be run with edge computing, the models must be designed based on the resources available on the device. In this work, we have evaluated different multilayer perceptron models for microcontrollers based on a number of different experiments. The machine learning models have been designed to detect road surfaces. The goal has been to identify how different parameters affect the machine learning systems. We have tried to maximize the performance and minimize the memory allocation of the models. The models have been designed to run on a microcontroller on the edge. The data was collected using an accelerometer integrated in a microcontroller mounted on a bicycle. The study evaluates two different machine learning systems that were developed in a previous thesis. The main focus of the work has been to create algorithms for detecting road surfaces. The data collection was done with a microcontroller equipped with an accelerometer mounted on a bicycle. One of the systems succeeds in achieving an accuracy of 93.1\% for the classification of 3 road surfaces. The work also evaluates how much physical memory is required by the various machine learning systems. The systems required between 1.78kB and 5,71kB of physical memory.
10
FASSARELA, M. S. "Treinamento de Redes Perceptron Usando Janelas Dinâmicas." Universidade Federal do Espírito Santo, 2009. http://repositorio.ufes.br/handle/10/9587.
Full textAbstract:
Made available in DSpace on 2018-08-02T00:00:48Z (GMT). No. of bitstreams: 1
tese_2871_DissertacaoMestradoMarceloSouzaFassarella.pdf: 4412674 bytes, checksum: b98757b8830dc327c5ca5578387c8eaa (MD5)
Previous issue date: 2009-12-21Neste trabalho apresentamos as redes neurais e o problema envolvendo o dilema bias-variância. Propomos o método da Janela a ser inserido no treinamento de redes supervisionadas com conjuntos de dados ruidosos. O método possui uma característica intrínseca de função regularizadora, já que procura eliminar ruídos durante a etapa de treinamento, reduzindo a in uência destes no ajuste dos pesos da rede. Implementamos e analisamos o método nas redes lógicas adaptivas (ALN) e nas redes perceptrons de múltiplas camadas (MLP). Por último, testamos a rede em aplicações de aproximação de funções, ltragem adaptiva e previsão de séries temporais.
11
Samuel, Nikhil J. "Identification of Uniform Class Regions using Perceptron Training." University of Cincinnati / OhioLINK, 2015. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1439307102.
Full text
12
Marchi, Rodrigo Andreoli de. "Aplicação do perceptron de múltiplas camadas no controle direto de potência do gerador de indução duplamente alimentado." [s.n.], 2011. http://repositorio.unicamp.br/jspui/handle/REPOSIP/258919.
Full textAbstract:
Orientadores: Edson Bim, Fernando José Von ZubenDissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação
Made available in DSpace on 2018-08-18T01:48:41Z (GMT). No. of bitstreams: 1 Marchi_RodrigoAndreolide_M.pdf: 5160130 bytes, checksum: 8e043b48f5cd0e86cc078b3adc27f74b (MD5) Previous issue date: 2011
Resumo: Neste trabalho é apresentada a estratégia de Controle Direto de Potência para o Gerador de Indução Duplamente Alimentado utilizando um controlador Perceptron de Múltiplas Camadas. O controlador tem a função de gerar os sinais das componentes de eixo direto e quadratura da tensão do rotor, sem a necessidade de controladores de corrente. A estratégia de controle apresentada permite operar o conversor de potência, conectado aos terminais do rotor, com frequência de chaveamento constante. A rede neural foi treinada off-line, a partir de um algoritmo de otimização de segunda ordem baseado no gradiente conjugado estendido, utilizando um conjunto de amostras obtido por meio da simulação digital de uma máquina de rotor bobinado de potência igual a 2 MW. Resultados de simulação digital com os dados dessa máquina, operando no modo gerador e com dupla alimentação, são apresentados para vários valores de potência ativa e reativa, e para velocidades fixas e variáveis, compreendidas na faixa de -15% a +15% da velocidade síncrona. Com o controlador implementado por uma rede neural artificial e treinada para uma máquina de 2 MW, testes de simulação digital e experimentais para uma máquina de 2,2 kW, operando na velocidade subsíncrona, são apresentados para validar a proposta
Abstract: This work presents a direct power control strategy for the doubly fed induction generator using a controller artificial neural networks, more specifically a multilayer perceptron. The controller has the role of generating the direct and quadrature-axis component signals of the rotor voltage, without the need of current controllers. The proposed control strategy allows to operate the converter, connected to the rotor terminals, with a fixed switching frequency. The multilayer perceptron was subject to an off-line training procedure using a second order algorithm based on an extend version of the conjugate gradient algorithm, using a set of samples produced by a 2 MW machine's digital simulation. Results of digital simulation for this machine are presented for several values of active and reactive power, with the generator operating on fixed and variable speed, in the range of -15% and +15% of the synchronous speed, considering the parameters of 2 MW machine. With the artificial neural network controller designed for this machine, digital simulation tests and experimental tests for a 2,2 kW machine, operating in a sub-synchronous speed, arc presented to validate the proposal
Mestrado
Energia Eletrica
Mestre em Engenharia Elétrica
13
Conan-Guez, Brieuc. "Modélisation supervisée de données fonctionnelles par perceptron multi-couches." Phd thesis, Université Paris Dauphine - Paris IX, 2002. http://tel.archives-ouvertes.fr/tel-00178892.
Full textAbstract:
L'Analyse de Données Fonctionnelles est une extension de l'analyse de données traditionnelles à des individus décrits par des fonctions. Le travail présenté ici s'inscrit pleinement dans ce courant, et tente de faire la jonction entre le domaine de la statistique fonctionnelle, et celui des techniques "neuronales" classiques. L'extension du perceptron multi-couches (PMC) à des espaces fonctionnels, proposé dans ce travail, apporte une réponse naturelle au traitement d'individus de type fonctions. Deux approches distinctes sont ici présentées : une approche par traitement direct des fonctions d'entrée et une approche par projection sur une base topologique de l'espace fonctionnel considéré (méthode classique en Analyse de Données Fonctionnelles). Pour chacune de ces deux méthodes, on montre dans un premier temps que le modèle est un approximateur universel, i.e. que toute fonction continue définie sur un compact d'un espace fonctionnel peut être approchée arbitrairement bien par un PMC fonctionnel. Dans un deuxième temps, on s'intéresse aux propriétés de consistance de l'estimateur fonctionnel. L'originalité de ce résultat vient du fait que non seulement l'estimation s'effectue sur un nombre fini d'individus (les fonctions observées), mais que de plus chacune de ces fonctions n'est connue qu'en un nombre fini de points d'observation (discrétisation). Un point important à noter est que ce résultat s'appuie sur une modélisation aléatoire du design des fonctions d'entrée. Enfin, on montre que le modèle peut encore être adapté afin d'obtenir une réponse fonctionnelle, ce qui autorise le traitement de processus fonctionnels à temps discret. L'approximation universelle et la consistance de l'estimateur (dans le cas i.i.d) sont encore vérifiées.
14
Zheng, Gonghui. "Design and evaluation of a multi-output-layer perceptron." Thesis, University of Ulster, 1996. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.338195.
Full text
15
Coelho, Maurício Archanjo Nunes. "Uma abordagem de predição estruturada baseada no modelo perceptron." Universidade Federal de Juiz de Fora (UFJF), 2015. https://repositorio.ufjf.br/jspui/handle/ufjf/3552.
Full textAbstract:
Submitted by Renata Lopes (renatasil82@gmail.com) on 2017-03-06T17:58:43Z
No. of bitstreams: 1
mauricioarchanjonunescoelho.pdf: 10124655 bytes, checksum: 549fa53eba76e81b76ddcbce12c97e55 (MD5)Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-03-06T20:26:43Z (GMT) No. of bitstreams: 1 mauricioarchanjonunescoelho.pdf: 10124655 bytes, checksum: 549fa53eba76e81b76ddcbce12c97e55 (MD5)
Made available in DSpace on 2017-03-06T20:26:44Z (GMT). No. of bitstreams: 1 mauricioarchanjonunescoelho.pdf: 10124655 bytes, checksum: 549fa53eba76e81b76ddcbce12c97e55 (MD5) Previous issue date: 2015-06-25
CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
A teoria sobre aprendizado supervisionado tem avançado significativamente nas últimas décadas. Diversos métodos são largamente utilizados para resoluções dos mais variados problemas, citando alguns: sistemas especialistas para obter respostas to tipo verdadeiro/ falso, o modelo Perceptron para separação de classes, Máquina de Vetores Suportes (SVMs) e o Algoritmo de Margem Incremental (IMA) no intuito de aumentar a margem de separação, suas versões multi-classe, bem como as redes neurais artificiais, que apresentam possibilidades de entradas relativamente complexas. Porém, como resolver tarefas que exigem respostas tão complexas quanto as perguntas? Tais respostas podem consistir em várias decisões inter-relacionadas que devem ser ponderadas uma a uma para se chegar a uma solução satisfatória e globalmente consistente. Será visto no decorrer do trabalho que existem problemas de relevante interesse que apresentam estes requisitos. Uma questão que naturalmente surge é a necessidade de se lidar com a explosão combinatória das possíveis soluções. Uma alternativa encontrada apresenta-se através da construção de modelos que compactam e capturam determinadas propriedades estruturais do problema: correlações sequenciais, restrições temporais, espaciais, etc. Tais modelos, chamados de estruturados, incluem, entre outros, modelos gráficos, tais como redes de Markov e problemas de otimização combinatória, como matchings ponderados, cortes de grafos e agrupamentos de dados com padrões de similaridade e correlação. Este trabalho formula, apresenta e discute estratégias on-line eficientes para predição estruturada baseadas no princípio de separação de classes derivados do modelo Perceptron e define um conjunto de algoritmos de aprendizado supervisionado eficientes quando comparados com outras abordagens. São também realizadas e descritas duas aplicações experimentais a saber: inferência dos custos das diversas características relevantes para a realização de buscas em mapas variados e a inferência dos parâmetros geradores dos grafos de Markov. Estas aplicações têm caráter prático, enfatizando a importância da abordagem proposta.
The theory of supervised learning has significantly advanced in recent decades. Several methods are widely used for solutions of many problems, such as expert systems for answers to true/false, Support Vector Machine (SVM) and Incremental Margin Algorithm (IMA). In order to increase the margin of separation, as well as its multi-class versions, in addition to the artificial neural networks which allow complex input data. But how to solve tasks that require answers as complex as the questions? Such responses may consist of several interrelated decisions to be considered one by one to arrive at a satisfactory and globally consistent solution. Will be seen throughout the thesis, that there are problems of relevant interest represented by these requirements. One question that naturally arises is the need to deal with the exponential explosion of possible answers. As a alternative, we have found through the construction of models that compress and capture certain structural properties of the problem: sequential correlations, temporal constraints, space, etc. These structured models include, among others, graphical models, such as Markov networks and combinatorial optimization problems, such as weighted matchings, graph cuts and data clusters with similarity and correlation patterns. This thesis formulates, presents and discusses efficient online strategies for structured prediction based on the principle of separation of classes, derived from the Perceptron and defines a set of efficient supervised learning algorithms compared to other approaches. Also are performed and described two experimental applications: the costs prediction of relevant features on maps and the prediction of the probabilistic parameters for the generating Markov graphs. These applications emphasize the importance of the proposed approach.
16
Ignatavičienė, Ieva. "Tiesioginio sklidimo neuroninių tinklų sistemų lyginamoji analizė." Master's thesis, Lithuanian Academic Libraries Network (LABT), 2012. http://vddb.laba.lt/obj/LT-eLABa-0001:E.02~2012~D_20120801_133809-03141.
Full textAbstract:
Pagrindinis darbo tikslas – atlikti kelių tiesioginio sklidimo neuroninių tinklų sistemų lyginamąją analizę siekiant įvertinti jų funkcionalumą.
Šiame darbe apžvelgiama: biologinio ir dirbtinio neuronų modeliai, neuroninių tinklų klasifikacija pagal jungimo konstrukciją (tiesioginio sklidimo ir rekurentiniai neuroniniai tinklai), dirbtinių neuroninių tinklų mokymo strategijos (mokymas su mokytoju, mokymas be mokytojo, hibridinis mokymas). Analizuojami pagrindiniai tiesioginio sklidimo neuroninių tinklų metodai: vienasluoksnis perceptronas, daugiasluoksnis perceptronas realizuotas „klaidos skleidimo atgal” algoritmu, radialinių bazinių funkcijų neuroninis tinklas.
Buvo nagrinėjama 14 skirtingų tiesioginio sklidimo neuroninių tinklų sistemos. Programos buvo suklasifikuotos pagal kainą, tiesioginio sklidimo neuroninių tinklo mokymo metodų taikymą, galimybę vartotojui keisti parametrus prieš apmokant tinklą ir techninį programos įvertinimą. Programos buvo įvertintos dešimtbalėje vertinimo sistemoje pagal mokymo metodų įvairumą, parametrų keitimo galimybes, programos stabilumą, kokybę, bei kainos ir kokybės santykį. Aukščiausiu balu įvertinta „Matlab” programa (10 balų), o prasčiausiai – „Sharky NN” (2 balai).
Detalesnei analizei pasirinktos keturios programos („Matlab“, „DTREG“, „PathFinder“, „Cortex“), kurios buvo įvertintos aukščiausiais balais, galėjo apmokyti tiesioginio sklidimo neuroninį tinklą daugiasluoksnio perceptrono metodu ir bent dvi radialinių bazinių funkcijų... [toliau žr. visą tekstą]The main aim – to perform a comparative analysis of several feedforward neural system networks in order to identify its functionality. The work presents both: biological and artificial neural models, also classification of neural networks, according to connections’ construction (of feedforward and recurrent neural networks), studying strategies of artificial neural networks (with a trainer, without a trainer, hybrid). The main methods of feedforward neural networks: one-layer perceptron, multilayer perceptron, implemented upon “error feedback” algorithm, also a neural network of radial base functions have been considered. The work has included 14 different feedforward neural system networks, classified according its price, application of study methods of feedforward neural networks, also a customer’s possibility to change parameters before paying for the network and a technical evaluation of a program. The programs have been evaluated from 1 point to 10 points according to the following: variety of training systems, possibility to change parameters, stability, quality and ratio of price and quality. The highest evaluation has been awarded to “Matlab” (10 points), the lowest – to “Sharky NN” (2 points). Four programs (”Matlab“, “DTREG“, “PathFinder“,”Cortex“) have been selected for a detail analysis. The best evaluated programs have been able to train feedforward neural networks using multilayer perceptron method, also at least two radial base function networks. “Matlab“ and... [to full text]
17
GOMES, Gecynalda Soares da Silva. "Novas funções de ativação em redes neurais artificiais multilayer perceptron." Universidade Federal de Pernambuco, 2010. https://repositorio.ufpe.br/handle/123456789/1757.
Full textAbstract:
Made available in DSpace on 2014-06-12T15:52:11Z (GMT). No. of bitstreams: 2
arquivo3194_1.pdf: 1782444 bytes, checksum: 1982844f90df3787391d8faa431cde16 (MD5)
license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5)
Previous issue date: 2010Conselho Nacional de Desenvolvimento Científico e Tecnológico
Em redes neurais artificiais (RNAs), as funções de ativação mais comumente usadas são a função sigmóide logística e a função tangente hiperbólica, dependendo das características dos dados. Entretanto, a escolha da função de ativação pode influenciar fortemente o desempenho e a complexidade da rede neural. Neste trabalho, com o objetivo de melhorar o desempenho dos modelos de redes neurais, propomos o uso de novas funções de ativação no processamento das unidades da rede neural. Aqui, as funções não-lineares implementadas são as inversas das funções de ligação usadas em modelos de regressão binomial, essas funções são: complemento log-log, probit, log-log e Aranda, sendo que esta última função apresenta um parâmetro livre e é baseada na família de transformações Aranda-Ordaz. Uma avaliação dos resultados do poder de predição com estas novas funções através de simulação Monte Carlo é apresentada. Além disso, foram realizados diversos experimentos com aproximação de funções contínuas e arbitrárias, com regressão e com previsão de séries temporais. Na utilização da função de ativação com parâmetro livre, duas metodologias foram usadas para a escolha do parâmetro livre, l . A primeira foi baseada em um procedimento semelhante ao de busca em linha (line search). A segunda foi usada uma metodologia para a otimização global dessa família de funções de ativação com parâmetro livre e dos pesos das conexões entre as unidades de processamento da rede neural. A ideia central é otimizar simultaneamente os pesos e a função de ativação usada em uma rede multilayer perceptron (MLP), através de uma abordagem que combina as vantagens de simulated annealing, de tabu search e de um algoritmo de aprendizagem local. As redes utilizadas para realizar esses experimentos foram treinadas através dos seguintes algoritmos de aprendizagem: backpropagation (BP), backpropagation com momentum (BPM), backpropagation baseado no gradiente conjugado com atualizações Fletcher-Reeves (CGF) e Levenberg-Marquardt (LM)
18
Lont, Jerzy B. "Analog CMOS implementation of a multi-layer perceptron with nonlinear synapses /." [S.l.] : [s.n.], 1993. http://e-collection.ethbib.ethz.ch/show?type=diss&nr=10244.
Full text
19
Vaughn, Marilyn Lougher. "Interpretation and knowledge discovery from the multi-layer perceptron neural network." Thesis, Cranfield University, 2005. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.427505.
Full text
20
Siu, Sammy. "Non-linear adaptive equalization based on a multi-layer perceptron architecture." Thesis, University of Edinburgh, 1991. http://hdl.handle.net/1842/11916.
Full textAbstract:
The subject of this thesis is the original study of the application of the multi-layer perceptron architecture to channel equalization in digital communications systems. Both theoretical analyses and simulations were performed to explore the performance of the perceptron-based equalizer (including the decision feedback equalizer). Topics covered include the factors that affect performance of the structures including, the parameters (learning gain and momentum parameter) in the learning algorithm, the network topology (input dimension, number of neurons and the number of hidden layers), and the power metrics on the error cost function. Based on the geometric hyperplane analysis of the multi-layer perceptron, the results offer valuable insight into the properties and complexity of the network. Comparisons of the bit error rate performance and the dynamic behaviour of the decision boundary of the perceptron-based equalizer with both the optimal non-linear equalizer and the optimal linear equalizer are provided. Through comparisons, some asymptotic results for the performance in the perceptron-based equalizer are obtained. Furthermore, a comparison of the performance of the perceptron-based equalizer (including the decision feedback equalizer) with the least mean squares linear transversal equalizer (including decision feedback equalizer) indicates that the former offers significant reduction in the bit error rate. This is because it has the ability to form highly nonlinear decision regions, in contrast with the linear equalizer which only forms linear decision regions. The linearity of the decision regions limits the performance of the conventional linear equalizer.
21
Birkmire, Brian Michael. "Weapon Engagement Zone Maximum Launch Range Approximation using a Multilayer Perceptron." Wright State University / OhioLINK, 2011. http://rave.ohiolink.edu/etdc/view?acc_num=wright1313763379.
Full text
22
Acuna, David A. Elizondo. "The recursive deterministic perceptron and topology reduction strategies for neural networks." Université Louis Pasteur (Strasbourg) (1971-2008), 1997. http://www.theses.fr/1997STR13001.
Full textAbstract:
Les strategies de reduction de la topologie des reseaux de neurones peuvent potentiellement offrir des avantages en termes de temps d'apprentissage, d'utilisation, de capacite de generalisation, de reduction des besoins materiels, ou comme etant plus proches du modele biologique. Apres avoir presente un etat de l'art des differentes methodes existantes pour developper des reseaux des neurones partiellement connectes, nous proposons quelques nouvelles methodes pour reduir le nombre de neurones intermediaires dans une topologie de reseaux neuronal. Ces methodes sont basees sur la notion de connexions d'ordre superieur. Un nouvel algorithme pour tester la separabilite lineaire et, d'autre part, une borne superieure de convergence pour l'algorithme d'apprentissage du perceptron sont donnes. Nous presentons une generalisation du reseau neuronal du perceptron, que nous nommons perceptron deterministe recursif (rdp) qui permet dans tous les cas de separer deux classes, de facon deterministe (meme si les deux classes ne sont pas directement lineairement separables). Cette generalisation est basee sur l'augmentation de la dimension du vecteur d'entree, laquelle produit plus de degres de liberte. Nous proposons une nouvelle notion de separabilite lineaire pour m classes et montrons comment generaliser le rdp a m classes en utilisant cette nouvelle notion
23
Satravaha, Nuttavudh. "Tone classification of syllable-segmented Thai speech based on multilayer perceptron." Morgantown, W. Va. : [West Virginia University Libraries], 2002. http://etd.wvu.edu/templates/showETD.cfm?recnum=2280.
Full textAbstract:
Thesis (Ph. D.)--West Virginia University, 2002.Title from document title page. Document formatted into pages; contains v, 130 p. : ill. (some col.). Vita. Includes abstract. Includes bibliographical references (p. 107-118).
24
Sevelev, Maxime. "Phase diagram, jamming and glass transitions in the non-convex perceptron." Thesis, Université Paris-Saclay (ComUE), 2017. http://www.theses.fr/2017SACLS331.
Full textAbstract:
Cette thèse de doctorat traite du « modèle de perceptron sphérique », un modèle simple et exactement soluble qui présente un comportement visqueux et d'encombrement qui a été généralisé aux valeurs négatives du paramètre de produit scalaire κ. Le problème classique d'apprentissage par machine qui consiste en la classification des motifs aléatoires par le perceptron fait partie des problèmes de satisfaction des contraintes (PSC) convexes. Même quand le « paramètre de stabilité » du modèle κ devient négatif, le problème reste toujours correctement posé et peut être interprété comme le problème de placement des particules sur une sphère N-dimensionnelle en évitant les obstacles placés au hasard. Dans ce cas, le PSC correspondant n'est pas convexe. Cette thèse étudie le problème en détail dans le domaine non convexe. Une étude systématique est rendue possible en faisant correspondre à un problème de satisfaction de contraintes un problème d'optimisation sur le même support, mais doté d'un Hamiltonien (fonction de coût) qui mesure les violations des contraintes en fonction de la configuration du système. Le lien entre le PSC aléatoire et la phénoménologie vitreuse en physique est bien connue et a été explorée en détail pour les modèles à variables discrètes. La présence de variables continues dans le modèle de perceptron (sphérique) nous permet de dévoiler, en PSC aléatoire, la transition caractéristique SAT/UNSAT où le système subit une transition du régime satisfaisable (dans lequel l'état fondamental possède une énergie nulle) à celui insatisfaisable (dans lequel l'état fondamental possède une énergie positive). Cette transition de phase peut également être interprétée comme une transition d'encombrement similaire à celles démontrées par les modèles des sphères sans friction. La simplicité du modèle étudié permet de trouver exactement son diagramme de phase à température zéro en fonction des deux paramètres de contrôle: la densité des obstacles et leur taille. Ainsi identifiée, la transition d'encombrement est complètement caractérisée dans le présent document. Sont également étudiées en détail de diverses phases vitreuses de caractère stable et marginalThis thesis treats the «spherical perceptron model», a simple exactly solvable model for glassy behavior and jamming suitably generalized to negative values of scalar product parameter κ. The classical machine-learning problem of random pattern classification by the perceptron is a convex constraint satisfaction problem (CSP). Even when the «stability parameter» κ of the model becomes negative, the problem still make sense and can be interpreted as the problem of particles on an N-dimensional sphere trying to avoid randomly placed obstacles. In this case, the corresponding CSP is non-convex. This thesis studies the problem in detail in the non-convex domain. Systematic study is made possible by assigning to a constraint satisfaction problem its corresponding optimization version endowed with a Hamiltonian function (cost function) quantifying the violations of the constraints, as a function of the system's configuration. The connection between random CSP and glassy phenomenology in physics is well known and has been explored in detail for models with discrete variables. The presence of continuous variables in the (spherical) perceptron model enables us to unveil, in random CSP, the characteristic SAT/UNSAT transition where the system transits from the satisfiable regime (where the ground state has zero energy) to the unsatisfiable one (where the ground state energy is positive). This phase transition can also be interpreted as a jamming transition similar to the one that exhibit models with frictionless spheres. The simplicity of the considered model allows the exact determination of the zero temperature phase diagram as a function of the control parameters: the density of obstacles and their size. In the present thesis, the jamming transition thus identified is completely characterized and several glass phases of stable and marginal character are studied in detail
25
Foxall, Robert John. "Likelihood analysis of the multi-layer perceptron and related latent variable models." Thesis, University of Newcastle Upon Tyne, 2001. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.327211.
Full text
26
Barbato, Daniela Maria Lemos. "O efeito das lesões nas capacidades de memorização e generalização de um perceptron." Universidade de São Paulo, 1993. http://www.teses.usp.br/teses/disponiveis/54/54131/tde-05092008-144618/.
Full textAbstract:
Perceptrons são redes neurais sem retroalimentação onde os neurônios estão dispostos em camadas. O perceptron considerado neste trabalho consiste de uma camada de N neurônios sensores Si = ±1; i = 1, , N ligados a um neurônio motor δ através das conexões sinápticas (pesos) Wi; i = 1, ..., N cujos valores restringimos a ±1. Utilizando o formalismo de Mecânica Estatística desenvolvido por Gardner (1988), estudamos os efeitos de eliminarmos uma fração de conexões sinápticas (diluição ) nas capacidades de memorização e generalização da rede neural descrita acima. Consideramos também o efeito de ruído atuando durante o estágio de treinamento do perceptron. Consideramos dois tipos de diluição: diluição móvel na qual os pesos são cortados de maneira a minimizar o erro de treinamento e diluição fixa na qual os pesos são cortados aleatoriamente. A diluição móvel, que modela lesões em cérebro de pacientes muito jovens, pode melhorar a capacidade de memorização e, no caso da rede ser treinada com ruído, também pode melhorar a capacidade de generalização. Por outro lado, a diluição fixa, que modela lesões em cérebros de pacientes adultos, sempre degrada o desempenho da rede, sendo seu principal efeito introduzir um ruído efetivo nos exemplos de treinamento.Perceptrons are layered, feed-forward neural networks. In this work we consider a per-ceptron composed of one input layer with N sensor neurons Si = ±1; i = 1, ... , N which are connected to a single motor neuron δ through the synaptic weights Wj; i = 1, ... , N, which are constrained to take on the values ±1 only. Using the Statistical Mechanics formalism developed by Gardner (1988), we study the effects of eliminating a fraction of synaptic weights on the memorization and generalization capabilities of the neural network described above. We consider also the effects of noise acting during the perceptron training stage. We consider two types of dilution: annealed dilution, where the weights are cut so as to minimize the training error and quenched dilution, where the weights are cut randomly. The annealed dilution which models brain damage in very young patients can improve the memorization ability and, in the case of training with noise, it can also improve the generalization ability. On the other hand, the quenched dilution which models lesions on adult brains always degrades the performance of the network, its main effect being to introduce an effective noise in the training examples.
27
Souza, Francisco Ary Alves de. "An?lise de desempenho da rede neural artificial do tipo multilayer perceptron na era multicore." Universidade Federal do Rio Grande do Norte, 2012. http://repositorio.ufrn.br:8080/jspui/handle/123456789/15447.
Full textAbstract:
Made available in DSpace on 2014-12-17T14:56:07Z (GMT). No. of bitstreams: 1
FranciscoAAS_DISSERT.pdf: 1526658 bytes, checksum: 7ba5b80f03a10eaf25a4f9e6a4c91372 (MD5)
Previous issue date: 2012-08-07Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior
Artificial neural networks are usually applied to solve complex problems. In problems with more complexity, by increasing the number of layers and neurons, it is possible to achieve greater functional efficiency. Nevertheless, this leads to a greater computational effort. The response time is an important factor in the decision to use neural networks in some systems. Many argue that the computational cost is higher in the training period. However, this phase is held only once. Once the network trained, it is necessary to use the existing computational resources efficiently. In the multicore era, the problem boils down to efficient use of all available processing cores. However, it is necessary to consider the overhead of parallel computing. In this sense, this paper proposes a modular structure that proved to be more suitable for parallel implementations. It is proposed to parallelize the feedforward process of an RNA-type MLP, implemented with OpenMP on a shared memory computer architecture. The research consistes on testing and analizing execution times. Speedup, efficiency and parallel scalability are analyzed. In the proposed approach, by reducing the number of connections between remote neurons, the response time of the network decreases and, consequently, so does the total execution time. The time required for communication and synchronization is directly linked to the number of remote neurons in the network, and so it is necessary to investigate which one is the best distribution of remote connections
As redes neurais artificiais geralmente s?o aplicadas ? solu??o de problemas comple- xos. Em problemas com maior complexidade, ao aumentar o n?mero de camadas e de neur?nios, ? poss?vel conseguir uma maior efici?ncia funcional, por?m, isto acarreta em um maior esfor?o computacional. O tempo de resposta ? um fator importante na decis?o de us?-las em determinados sistemas. Muitos defendem que o maior custo computacional est? na fase de treinamento. Por?m, esta fase ? realizada apenas uma ?nica vez. J? trei- nada, ? necess?rio usar os recursos computacionais existentes de forma eficiente. Diante da era multicore esse problema se resume ? utiliza??o eficiente de todos os n?cleos de processamento dispon?veis. No entanto, ? necess?rio considerar a sobrecarga existente na computa??o paralela. Neste sentido, este trabalho prop?e uma estrutura modular que ? mais adequada para as implementa??es paralelas. Prop?e-se paralelizar o processo feed- forward (passo para frente) de uma RNA do tipo MLP, implementada com o OpenMP em uma arquitetura computacional de mem?ria compartilhada. A investiga??o dar-se-? com a realiza??o de testes e an?lises dos tempos de execu??o. A acelera??o, a efici?ncia e a es- calabilidade s?o analisados. Na proposta apresentada ? poss?vel perceber que, ao diminuir o n?mero de conex?es entre os neur?nios remotos, o tempo de resposta da rede diminui e por consequ?ncia diminui tamb?m o tempo total de execu??o. O tempo necess?rio para comunica??o e sincronismo est? diretamente ligado ao n?mero de neur?nios remotos da rede, sendo ent?o, necess?rio observar sua melhor distribui??o
28
Fischer, Manfred M., and Petra Staufer-Steinnocher. "Spectral Pattern Recognition by a Two-Layer Perceptron: Effects of Training Set Size." WU Vienna University of Economics and Business, 1996. http://epub.wu.ac.at/4160/1/WSG_DP_5596.pdf.
Full textAbstract:
Pattern recognition in urban areas is one of the most challenging issues in
classifying satellite remote sensing data. Parametric pixel-by-pixel classification
algorithms tend to perform poorly in this context. This is because urban areas
comprise a complex spatial assemblage of disparate land cover types - including
built structures, numerous vegetation types, bare soil and water bodies. Thus,
there is a need for more powerful spectral pattern recognition techniques,
utilizing pixel-by-pixel spectral information as the basis for automated urban
land cover detection. This paper adopts the multi-layer perceptron classifier
suggested and implemented in [5]. The objective of this study is to analyse the
performance and stability of this classifier - trained and tested for supervised
classification (8 a priori given land use classes) of a Landsat-5 TM image
(270 x 360 pixels) from the city of Vienna and its northern surroundings
- along with varying the training data set in the single-training-site case.
The performance is measured in terms of total classification, map user's and
map producer's accuracies. In addition, the stability with initial parameter
conditions, classification error matrices, and error curves are analysed in some
detail. (authors' abstract)Series: Discussion Papers of the Institute for Economic Geography and GIScience
29
D'Alimonte, Davide. "Multi layer perceptron neural network algorithms for ocean colour applications in coastal waters." Thesis, University of Southampton, 2004. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.401830.
Full text
30
AmÃncio, Luciana Barbosa. "PrevisÃo de recalques em fundaÃÃes profundas utilizando redes neurais artificiais do tipo perceptron." Universidade Federal do CearÃ, 2013. http://www.teses.ufc.br/tde_busca/arquivo.php?codArquivo=10726.
Full textAbstract:
nÃo hÃA previsÃo de recalque em fundaÃÃes profundas do tipo estacas hÃlice contÃnua, escavada e cravada metÃlica à o objeto principal desse estudo. O recalque à o deslocamento vertical para baixo que uma fundaÃÃo apresenta quando submetida a um determinado carregamento. A estimativa dos recalques em fundaÃÃes profundas pode ser feita utilizando-se diversas metodologias, dentre as quais os mÃtodos numÃricos e os teÃricos. Diferentes variÃveis influenciam os recalques ocorridos nas fundaÃÃes profundas do tipo estaca destacando-se as caracterÃsticas de resistÃncia e deformabilidade dos materiais envolvidos, a estratigrafia do solo de fundaÃÃo e a geometria do elemento estrutural de fundaÃÃo, dentre outras, configurando-se, portanto, como um problema multi-variado e de grande complexidade. Uma alternativa para a estimativa mais realista dos recalques em fundaÃÃes profundas consiste no emprego das redes neurais artificiais, que sÃo modelos que trabalham analogamente ao cÃrebro humano que tÃm, recentemente, contribuÃdo na resoluÃÃo de problemas complexos em diversas Ãreas da Engenharia Civil. Nessa pesquisa foram utilizadas redes neurais multicamadas alimentadas adiante (perceptron multi-camadas) para o desenvolvimento de um modelo de previsÃo de recalques em estacas, a partir de um treinamento supervisionado, que utiliza o algoritmo de retropropagaÃÃo do erro (error back propagation). Para o desenvolvimento do modelo foram coletados resultados de ensaios SPT e provas de carga estÃtica, e com auxÃlio do programa QNET2000 foram treinados e validados vÃrios modelos de redes neurais. ApÃs as anÃlises e comparaÃÃes entre os resultados de diferentes configuraÃÃes, constatou-se que as redes neurais artificiais foram capazes de entender o comportamento das fundaÃÃes profundas do tipo hÃlice contÃnua, cravada metÃlica e escavada no que tange a influÃncia das variÃveis de entrada consideradas para a estimativa dos recalques. AlÃm disto, constatou-se que os resultados obtidos pelo modelo desenvolvido permitem, dentre outras coisas, a definiÃÃo das cargas de trabalho e cargas limites na estaca. A arquitetura desse modelo à formada por 6 nÃs na camada de entrada, 20 neurÃnios distribuÃdos em 3 camadas ocultas, e 1 neurÃnio na camada de saÃda, correspondente ao recalque medido para a estaca. O processo de alteraÃÃo dos pesos sinÃpticos, na fase de validaÃÃo do modelo, com 4 milhÃes de iteraÃÃes resultou no maior coeficiente de correlaÃÃo entre os recalques estimados e os recalques medidos, que foi de 0,89, o qual pode ser considerado satisfatÃrio, em se tratando da previsÃo de um fenÃmeno complexo.
The settlement deep foundations preview of stakes continuous helix, metallic dug and stuck is the aim of this study. The settlement is a vertical down dislocation a foundation shows when it undergoes a determined charge. The settlements assessment in deep foundations can be done using several methods as, for instance, the numerical and the theoretical ones. Different variables influence the settlements occurred in foundations of the stake kind which can be detached, among them, the characteristics of resistance and deformation of the involved material, the stratigraphy of the foundation ground and the geometry of the foundationâs structural element manifesting, thus, a multi-diverse and high-complex problem. An alternative to a more realistic assessment of the settlements in deep foundations consists in the application of the artificial neural networks, models that work analogically in the human brain which have been recently contributing to the resolution of complex problems in different areas of Civil Engineering. In this research, multi-marked neural networks were used, fed ahead (perceptron multi-layer) to develop a preview model of settlements in stakes, since a managed training which uses the error back propagation algorithm. To the development of the model, SPT experiments and static charge testsâ results were collected and, with the help of QNET 2000 program, several neural network models were tested and validated. After the analysis and comparison of the different configurationsâ results, it was verified that the artificial neural networks were able to understand the deep foundations behavior, continuous helix, metallic dug and stuck kind concerned to the influence of entrance variables considered to the settlements assessment. Furthermore, the results obtained by the developed model allow, through other factors, the definition of work charges and limit charges on the stake. The architecture of this model is formed by 6 knots in the entrance layer, 20 neurons distributed in 3 hidden layers and 1 neuron in the exit layer, corresponding to the measured settlements to the stake. The change process of the synaptic heights, in the modelâs validation stage, with 4 million iterations, resulted in the bigger correlation coefficient between the assessed and the measured settlements (0.89), which is satisfactory regarding the preview of a complex phenomenon.
31
Manesco, Luis Fernando. "Modelagem de um processo fermentativo por rede Perceptron multicamadas com atraso de tempo." Universidade de São Paulo, 1996. http://www.teses.usp.br/teses/disponiveis/18/18133/tde-22012018-103016/.
Full textAbstract:
A utilização de Redes Neurais Artificias para fins de identificação e controle de sistemas dinâmicos têm recebido atenção especial de muitos pesquisadores, principalmente no que se refere a sistemas não lineares. Neste trabalho é apresentado um estudo sobre a utilização de um tipo em particular de Rede Neural Artificial, uma Perceptron Multicamadas com Atraso de Tempo, na estimação de estados da etapa fermentativa do processo de Reichstein para produção de vitamina C. A aplicação de Redes Neurais Artificiais a este processo pode ser justificada pela existência de problemas associados à esta etapa, como variáveis de estado não mensuráveis e com incertezas de medida e não linearidade do processo fermentativo, além da dificuldade em se obter um modelo convencional que contemple todas as fases do processo. É estudado também a eficácia do algoritmo de Levenberg-Marquadt, na aceleração do treinamento da Rede Neural Artificial, além de uma comparação do desempenho de estimação de estados das Redes Neurais Artificiais estudadas com o filtro estendido de Kalman, baseado em um modelo não estruturado do processo fermentativo. A análise do desempenho das Redes Neurais Artificiais estudadas é avaliada em termos de uma figura de mérito baseada no erro médio quadrático sendo feitas considerações quanto ao tipo da função de ativação e o número de unidades da camada oculta. Os dados utilizados para treinamento e avaliação da Redes Neurais Artificiais foram obtidos de um conjunto de ensaios interpolados para o intervalo de amostragem desejado.ldentification and Control of dynamic systems using Artificial Neural Networks has been widely investigated by many researchers in the last few years, with special attention to the application of these in nonlinear systems. ls this works, a study on the utilization of a particular type of Artificial Neural Networks, a Time Delay Multi Layer Perceptron, in the state estimation of the fermentative phase of the Reichstein process of the C vitamin production. The use of Artificial Neural Networks can be justified by the presence of problems, such as uncertain and unmeasurable state variables and process non-linearity, and by the fact that a conventional model that works on all phases of the fermentative processes is very difficult to obtain. The efficiency of the Levenberg Marquadt algorithm on the acceleration of the training process is also studied. Also, a comparison is performed between the studied Artificial Neural Networks and an extended Kalman filter based on a non-structured model for this fermentative process. The analysis of lhe Artificial Neural Networks is carried out using lhe mean square errors taking into consideration lhe activation function and the number of units presents in the hidden layer. A set of batch experimental runs, interpolated to the desired time interval, is used for training and validating the Artificial Neural Networks.
32
Amancio, Luciana Barbosa. "Previsão de recalques em fundações profundas utilizando redes neurais artificiais do tipo Perceptron." reponame:Repositório Institucional da UFC, 2013. http://www.repositorio.ufc.br/handle/riufc/7980.
Full textAbstract:
AMANCIO, L. B. Previsão de recalques em fundações profundas utilizando redes neurais artificiais do tipo Perceptron. 2013. 90 f. Dissertação (Mestrado em Engenharia Civil: Geotecnia) - Centro de Tecnologia, Universidade Federal do Ceará, Fortaleza, 2013.Submitted by Marlene Sousa (mmarlene@ufc.br) on 2014-04-04T17:10:28Z No. of bitstreams: 1 2013_dis_lbamancio.pdf: 1795121 bytes, checksum: 22a58885fe6cc9e62c41ea27efa25d87 (MD5)
Approved for entry into archive by Marlene Sousa(mmarlene@ufc.br) on 2014-05-06T17:18:03Z (GMT) No. of bitstreams: 1 2013_dis_lbamancio.pdf: 1795121 bytes, checksum: 22a58885fe6cc9e62c41ea27efa25d87 (MD5)
Made available in DSpace on 2014-05-06T17:18:03Z (GMT). No. of bitstreams: 1 2013_dis_lbamancio.pdf: 1795121 bytes, checksum: 22a58885fe6cc9e62c41ea27efa25d87 (MD5) Previous issue date: 2013-08-16
The settlement deep foundations preview of stakes continuous helix, metallic dug and stuck is the aim of this study. The settlement is a vertical down dislocation a foundation shows when it undergoes a determined charge. The settlements assessment in deep foundations can be done using several methods as, for instance, the numerical and the theoretical ones. Different variables influence the settlements occurred in foundations of the stake kind which can be detached, among them, the characteristics of resistance and deformation of the involved material, the stratigraphy of the foundation ground and the geometry of the foundation’s structural element manifesting, thus, a multi-diverse and high-complex problem. An alternative to a more realistic assessment of the settlements in deep foundations consists in the application of the artificial neural networks, models that work analogically in the human brain which have been recently contributing to the resolution of complex problems in different areas of Civil Engineering. In this research, multi-marked neural networks were used, fed ahead (perceptron multi-layer) to develop a preview model of settlements in stakes, since a managed training which uses the error back propagation algorithm. To the development of the model, SPT experiments and static charge tests’ results were collected and, with the help of QNET 2000 program, several neural network models were tested and validated. After the analysis and comparison of the different configurations’ results, it was verified that the artificial neural networks were able to understand the deep foundations behavior, continuous helix, metallic dug and stuck kind concerned to the influence of entrance variables considered to the settlements assessment. Furthermore, the results obtained by the developed model allow, through other factors, the definition of work charges and limit charges on the stake. The architecture of this model is formed by 6 knots in the entrance layer, 20 neurons distributed in 3 hidden layers and 1 neuron in the exit layer, corresponding to the measured settlements to the stake. The change process of the synaptic heights, in the model’s validation stage, with 4 million iterations, resulted in the bigger correlation coefficient between the assessed and the measured settlements (0.89), which is satisfactory regarding the preview of a complex phenomenon.
A previsão de recalque em fundações profundas do tipo estacas hélice contínua, escavada e cravada metálica é o objeto principal desse estudo. O recalque é o deslocamento vertical para baixo que uma fundação apresenta quando submetida a um determinado carregamento. A estimativa dos recalques em fundações profundas pode ser feita utilizando-se diversas metodologias, dentre as quais os métodos numéricos e os teóricos. Diferentes variáveis influenciam os recalques ocorridos nas fundações profundas do tipo estaca destacando-se as características de resistência e deformabilidade dos materiais envolvidos, a estratigrafia do solo de fundação e a geometria do elemento estrutural de fundação, dentre outras, configurando-se, portanto, como um problema multi-variado e de grande complexidade. Uma alternativa para a estimativa mais realista dos recalques em fundações profundas consiste no emprego das redes neurais artificiais, que são modelos que trabalham analogamente ao cérebro humano que têm, recentemente, contribuído na resolução de problemas complexos em diversas áreas da Engenharia Civil. Nessa pesquisa foram utilizadas redes neurais multicamadas alimentadas adiante (perceptron multi-camadas) para o desenvolvimento de um modelo de previsão de recalques em estacas, a partir de um treinamento supervisionado, que utiliza o algoritmo de retropropagação do erro (error back propagation). Para o desenvolvimento do modelo foram coletados resultados de ensaios SPT e provas de carga estática, e com auxílio do programa QNET2000 foram treinados e validados vários modelos de redes neurais. Após as análises e comparações entre os resultados de diferentes configurações, constatou-se que as redes neurais artificiais foram capazes de entender o comportamento das fundações profundas do tipo hélice contínua, cravada metálica e escavada no que tange a influência das variáveis de entrada consideradas para a estimativa dos recalques. Além disto, constatou-se que os resultados obtidos pelo modelo desenvolvido permitem, dentre outras coisas, a definição das cargas de trabalho e cargas limites na estaca. A arquitetura desse modelo é formada por 6 nós na camada de entrada, 20 neurônios distribuídos em 3 camadas ocultas, e 1 neurônio na camada de saída, correspondente ao recalque medido para a estaca. O processo de alteração dos pesos sinápticos, na fase de validação do modelo, com 4 milhões de iterações resultou no maior coeficiente de correlação entre os recalques estimados e os recalques medidos, que foi de 0,89, o qual pode ser considerado satisfatório, em se tratando da previsão de um fenômeno complexo.
33
Souza, Ana Paula de. "Coerência, modelo oculto de Markov e Perceptron de multi-camadas em imagética motora." Universidade Federal de Minas Gerais, 2010. http://hdl.handle.net/1843/BUOS-8CLJ44.
Full textAbstract:
Brain signals and the interpretation of their patterns provide a new modality of communication: the Brain Computer Interface (BCI). BCI can use scalp potential related movement imagination to activate drive devices, not depending on the brains normal output channels: peripheral nerves and muscles. Magnitude Squared Coherence has beenused to identify the event related potential in Electroencephalogram (EEG) signals. Moreover, techniques such as Hidden Markov Model (HMM) and Artificial Neural Network (ANN with Multilayer Perceptron MLP structure) have shown promising results in classification for BCI systems. Thus, this work aims to investigate classification using HMM and ANN using features from MSC in EEG signals, for the following events: spontaneous EEG; actual index finger movement; and imaginary movement of that finger. EEGs were recorded from three normal subjects from electrodes placed according to the International 10-20 System (1st record) and 10-10 System (2nd and 3rd record). EEG was divided into trials (M - 14 seconds each) synchronized with the event. Each trial wasdivided into six segments: spontaneous EEG; EEG during visualization of red LED (Light Emitting Diode) attention; EEG during visualization of yellow LED preparation for the event; EEG during the event; spontaneous EEG; and spontaneous EEG. MSC was calculated for 12 trials and afterwards, for the maximum trials existent in each electrode.In each segment the MSC was calculated for delta band (0.1 2.0 Hz), alpha band (8.0 13 Hz) and beta band (14 30 Hz), with M=12 trials and M = maximum number of trials. The frequency band that presented the highest MSC was used as observation in HMM and as input for RNA. The average accuracy rates in the classification using HMM for M = 12were 68.5 %, 66.5 % and 67.5 %, for subjects #1, #2 and #3, respectively. For maximum M, they were 73.0 %, 70.0 % and 56.5 %. When MLP was used for classification the results for 12 trials were 64.0 %, 75.5 % and 82.0 % and, using maximum M, the accuracy rates obtained were 79.5 %, 85.5 % and 88.5 %. These results showed that the MSC technique is an efficient tool for feature extraction in EEG recording during differentevents. With these features, it was possible to classify the EEG signals using HMM and RNA, the latter presenting the best performance in event classification.Sinais cerebrais e a interpretação de padrões nos mesmos propiciam uma nova modalidade de comunicação: a Interface Cérebro Máquina (ICM). A ICM pode utilizar a atividade elétrica advinda do córtex cerebral relacionada à imaginação motora para promover o acionamento de um dispositivo, sem que haja necessariamente a integridade das vias motoras. A Magnitude Quadrática da Coerência (MSC Magnitude Square Coherence) tem sido utilizada para identificar o potencial relacionado a eventos no sinal de eletroencefalograma (EEG). Além disso, para a classificação desses eventos e aplicações em ICM técnicas como os Modelos Ocultos de Markov (Hidden Markov Model - HMM) e as Redes Neurais Artificiais (RNA) têm se mostrado promissoras. Dessa forma, esta dissertação visa investigar a classificação com HMM e RNA (na estrutura Multilayer Peceptron - MLP) utilizando como característica a MSC do sinal EEG, para os eventos:espontâneo; movimento do dedo indicador da mão esquerda; e imaginação deste movimento. Sinais EEG de três voluntários durante os eventos foram coletados com eletrodos dispostos segundo o Sistema Internacional 10-20 (1ª coleta) e Sistema 10-10 (2ª e 3ª coletas). O EEG foi dividido em trechos (M 14 segundos cada) sincronizados com oevento. Cada trecho foi fragmentado em seis segmentos: EEG espontâneo; EEG do voluntário visualizando um LED (Light Emitting Diode) vermelho atenção; EEG visualizando um LED amarelo preparação para realizar o evento; EEG durante o evento; EEG espontâneo; EEG espontâneo. Em cada segmento avaliou-se a MSC da banda delta (0,1-2 Hz), banda alfa (8-13 Hz) e banda beta (14-30 Hz), com M=12 e M=máximo número de trechos. A faixa de freqüência que apresentou maior MSC foi adotada como observação a ser utilizada no HMM e como entrada para a RNA. Os índices médios de acerto na classificação com o HMM para M=12 foram 68,5 %, 66,5 % e 67,5 %, para os sujeitos #1, #2 e #3, respectivamente. Para M máximo foram de 73,0 %, 70,0 % e 56,5 %. Quando utilizada a MLP para a classificação os índices para M=12 foram 64,0 %, 75,5 % e 82,0 % e, para M máximo 79,5 %, 85,5 % e 88,5 %. Esses achados mostram que é possível fazer modelos e classificações utilizando HMM e RNA de atividades cerebraisrelacionadas a diferentes eventos, usando a MSC do sinal de EEG. A RNA mostrou melhor desempenho do que o HMM na classificação dos eventos.
34
Dufberg, Johan. "Automatisk dokumentklassificering med hjälp av maskininlärning." Thesis, Örebro universitet, Institutionen för naturvetenskap och teknik, 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:oru:diva-67228.
Full textAbstract:
Att manuellt hantera och klassificera stora mängder textdokument tar mycket tid och kräver mycket personal, att göra detta med hjälp av maskininlärning är för ändamålet ett alternativ. Det här arbetet önskar ge läsaren en grundläggande inblick i hur automatisk klassificering av texter fungerar, samt ge en lätt samanställning av några av de vanligt förekommande algoritmerna för ändamålet. De exempel som visas använder sig av artiklar på engelska om teknik- och finansnyheter, men arbetet har avstamp i frågan om mognadsgrad av tekniken för hantering av svenska officiella dokument. Första delen är den vetenskapliga bakgrund som den andra delen vilar på, här beskrivs flera algoritmer och tekniker som sedan används i praktiska exempel. Rapporten ämnar inte beskriva en färdig produkt, utan fungerar så som ”proof of concept” för textklassificeringens användning. Avslutningsvis diskuteras resultaten från de tester som gjorts, och en av slutsatserna är att när det finns tillräckligt med data kan en enkel klassificerare prestera nästan likvärdigt med en tekniskt sett mer utvecklad och komplex klassificerare. Relateras prestandan hos klassificeraren till tidsåtgången visar detta på att komplexa klassificerare kräver hårdvara med hög beräkningskapacitet och mycket minne för att vara gångbara.To manually handle and classify large quantities of text documents, takes a lot of time and demands a large staff, to use machine learning for this purpose is an alternative. This thesis aims to give the reader a fundamental insight in how automatic classification of texts work and give a quick overview of the most common algorithms used for this purpose. The examples that are shown uses news articles in English about tech and finance, but the thesis takes a start in the question about how mature the technique is for handling official Swedish documents. The first part is the scientific background on which the second part rests, here several algorithms and techniques are described which is used in practice later. The report does not aim to describe a product in any form but acts as a “proof of concept” for the use of text classification. Finally, the results from the tests are discussed, and one of the conclusions drawn is that when data is abundant a relatively simple classifier can perform close to equal to a technically more developed and complex classifier. If the performance of the classifier is related to the time taken this indicates that complex classifiers need hardware with high computational power and a fair bit of memory for the classifier to be viable.
35
Dhekne, Rucha P. "Machine Learning Techniques to Provide Quality of Service in Cognitive Radio Technology." University of Cincinnati / OhioLINK, 2009. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1258579803.
Full text
36
Östling, Robert. "Tagging a Morphologically Complex Language Using an Averaged Perceptron Tagger: The Case of Icelandic." Stockholms universitet, Avdelningen för datorlingvistik, 2013. http://urn.kb.se/resolve?urn=urn:nbn:se:su:diva-90304.
Full textAbstract:
In this paper, we experiment with using Stagger, an open-source implementation of an Averaged Perceptron tagger, to tag Icelandic, a morphologically complex language. By adding languagespecific linguistic features and using IceMorphy, an unknown word guesser, we obtain state-of- the-art tagging accuracy of 92.82%. Furthermore, by adding data from a morphological database, and word embeddings induced from an unannotated corpus, the accuracy increases to 93.84%. This is equivalent to an error reduction of 5.5%, compared to the previously best tagger for Icelandic, consisting of linguistic rules and a Hidden Markov Model.
37
Coelho, Maurício Archanjo Nunes. "Predição de dados estruturados utilizando a formulação Perceptron com aplicação em planejamento de caminhos." Universidade Federal de Juiz de Fora (UFJF), 2010. https://repositorio.ufjf.br/jspui/handle/ufjf/3597.
Full textAbstract:
Submitted by Renata Lopes (renatasil82@gmail.com) on 2017-03-07T15:27:21Z
No. of bitstreams: 1
mauricioarchanjonunescoelho.pdf: 2468130 bytes, checksum: 3f05daa8428e367942c4ad560b6375f2 (MD5)Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-03-10T12:19:13Z (GMT) No. of bitstreams: 1 mauricioarchanjonunescoelho.pdf: 2468130 bytes, checksum: 3f05daa8428e367942c4ad560b6375f2 (MD5)
Made available in DSpace on 2017-03-10T12:19:13Z (GMT). No. of bitstreams: 1 mauricioarchanjonunescoelho.pdf: 2468130 bytes, checksum: 3f05daa8428e367942c4ad560b6375f2 (MD5) Previous issue date: 2010-06-18
CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
O problema de planejamento de caminhos apresenta diversas subáreas, muitas das quais já extensamente abordadas na literatura. Uma dessas áreas em especial é a de determinação de caminhos, os algoritmos empregados para a solução deste problema dependem que os custos estipulados para os ambientes ou mapas sejam confiáveis. A dificuldade está justamente na definição dos custos referentes a cada tipo de área ou terreno nos mapas a serem examinados. Como se pode observar, o problema mencionado inclui a dificuldade em se determinar qual o custo de cada característica relevante presente no mapa, bem como os custos de suas possíveis combinações. A proposta deste trabalho é mostrar como é feita a predição desses custos em novos ambientes tendo como base a predição de dados estruturados definindo um aprendizado funcional entre domínios de entrada e saída, estruturados e arbitrários. O problema de aprendizado em questão é normalmente formulado como um problema de otimização convexa de máxima margem bastante similar a formulação de máquinas de vetores suporte multi-classe. Como técnica de solução realizou-se a implementação do algoritmo MMP (Maximum Margin Planning) (RATLIFF; BAGNELL; ZINKEVICH, 2006). Como contribuição, desenvolveu-se e implementou-se dois algoritmos alternativos, o primeiro denominado Perceptron Estruturado e o segundo Perceptron Estruturado com Margem, ambos os métodos de relaxação baseados na formulação do Perceptron. Os mesmos foram analisados e comparados. Posteriormente temos a exploração dos ambientes por um agente inteligente utilizando técnicas de aprendizado por reforço. Tornando todo o processo, desde a análise do ambiente e descoberta de custos, até sua exploração e planejamento do caminho, um completo processo de aprendizado.
The problem of path planning has several sub-areas, many of which are widely discussed in the literature. One of these areas in particular is the determination of paths, the algorithms used to solve this problem depend on the reliability of the estimated costs in the environments and maps. The difficulty is precisely the definition of costs for each type of area or land on the maps to be examined. As you can see, the problem mentioned includes the difficulty in determining what the cost of each relevant characteristic on the map, and the costs of their possible combinations. The purpose of this study is to show how the prediction of these costs is made into new environments based on the prediction of structured data by defining functional learning areas between input and output, structured and arbitrary. The problem of learning in question is usually formulated as a convex optimization problem of maximum margin very similar to the formulation of multiclass support vector machines. A solution technic was performed through implementation of the algorithm MMP (Maximum Margin Planning) (RATLIFF; BAGNELL; ZINKEVICH, 2006). As a contribution, two alternative algorithms were developed and implemented, the first named Structured Perceptron, and the second Structured Perceptron with Margin both methods of relaxation based formulation of the Perceptron. They were analyzed and compared. Posteriorly we have the exploitation of the environment by an intelligent agent using reinforcement learning techniques. This makes the whole process, from the environment analysis and discovery of cost to the exploitation and path planning, a complete learning process.
38
AraÃjo, Carla Beatriz Costa de. "AplicaÃÃo das redes neurais artificiais do tipo perceptron na estimativa de recalques em estacas." Universidade Federal do CearÃ, 2015. http://www.teses.ufc.br/tde_busca/arquivo.php?codArquivo=14556.
Full textAbstract:
A utilizaÃÃo das redes neurais artificiais (RNA) na estimativa de recalques em fundaÃÃes
profundas à comprovadamente uma ferramenta eficiente. Nos trabalhos de AmÃncio (2013) e
Silveira (2014), o emprego das RNA apresentou bons resultados para a previsÃo de recalques
em estacas hÃlices contÃnuas, estacas cravadas metÃlicas e estacas escavadas. PorÃm, algumas
estacas modeladas apresentaram comportamento muito distante dos resultados reais, onde os
resultados da modelagem indicaram aumentos bruscos na rigidez do sistema solo-estaca. Nesta
pesquisa, foi desenvolvido um modelo com uma rede neural do tipo perceptron multicamadas
de forma a melhorar o desempenho dos modelos de AmÃncio (2013) e Silveira (2014). Para
desenvolvimento do trabalho, inicialmente foram feitas anÃlises dos resultados de sondagens Ã
percussÃo do tipo SPT e provas de carga estÃticas das 199 estacas utilizadas no trabalho
apresentado por Silveira (2014), fazendo-se uma avaliaÃÃo da consistÃncia das informaÃÃes,
com o objetivo de ter um conjunto mais heterogÃneo e representativo. ApÃs a realizaÃÃo de
alteraÃÃes, chegou-se a um conjunto com 141 estacas, totalizando 1.320 exemplos do tipo
entrada-saÃda. Foram definidas como variÃveis de entrada do modelo: o tipo de estaca, o
comprimento da estaca, o diÃmetro da estaca, o nÃmero representativo dos valores de NSPT ao
longo do fuste da estaca (denominada NF), o NSPT na ponta da estaca, profundidade da camada
de influÃncia da carga em relaÃÃo a ponta da estaca, o fator representativo das camadas de solo
argiloso, o fator representativo das camadas de solo siltoso, o fator representativo das camadas
de solo arenoso e a carga aplicada. Foram estudadas quatro diferentes formas de cÃlculo da
variÃvel de entrada NF, sendo estas: soma, mÃdia, soma ponderada e mÃdia ponderada. Com as
variÃveis de entrada apresentadas foram trabalhados modelos onde a variÃvel de saÃda fosse o
recalque da fundaÃÃo profunda. A modelagem das RNA foi feita utilizando o programa QNET
2000, e foram realizados o treinamento e a validaÃÃo de diferentes arquiteturas. O modelo que
teve melhor desempenho apresentou coeficiente de correlaÃÃo entre os recalques reais e os
recalques modelados no treinamento de 0,99 e na validaÃÃo de 0,98. Os resultados obtidos
mostraram-se melhores que os de AmÃncio (2013) e Silveira (2014), que na fase de validaÃÃo,
apresentaram correlaÃÃes de 0,89 e 0,94 respectivamente. O modelo final deste trabalho possui
uma arquitetura formada por 10 nÃs na camada de entrada, 34 neurÃnios distribuÃdos ao longo
de quatro camadas ocultas e um neurÃnio na camada de saÃda (A:10-15-9-7-3-1), utilizando a
mÃdia para cÃlculo do nÃmero representativo dos valores de NSPT ao longo do fuste da estaca.use of artificial neural networks (ANN) in the estimation of settlements in foundations deep has proven an effective tool. The work of Amancio (2013) and Silveira (2014), the use of RNA showed good results for predicting settlements in continuous stakes propellers, metal piles driven and bored piles. However, some modeled stakes had far behavior of real results, where modeling results indicate sharp increases in stiffness soil-cutting system. In this research, it developed a model with a neural network of the multilayer perceptron to improve the performance of the models AmÃncio (2013) and Silveira (2014). To development work initially polls results of analyzes were made Percussion SPT and static load tests of 199 stakes used at work presented by Silveira (2014), making up an assessment of the consistency of the information, in order to have a more heterogeneous and the representative assembly. After conducting changes, has come up with a set with 141 stakes, totaling 1,320 examples of the type entrance exit. Were defined as model input variables: the type of pile, the length of the pile, the pile diameter, the number of representative values when NSPT Over stake stem (called NF), the NSPT on the edge of the pile, depth of the layer the influence of load relative to the cutting edge, the factor representative of the soil layers clay, the representative factor of silty soil layers, the representative factor of the layers sandy soil and the applied load. Four different ways of calculation have been studied in NF input variable, which are: sum, average, weighted sum and weighted average. With input variables presented were worked models where the output variable was the repression of deep foundation. The modeling of RNA was made using the QNET program 2000 and were carried out training and validation of different architectures. The model had better performance showed correlation coefficient between the actual settlements and settlements modeled in the training of 0.99 and 0.98 in the validation. The results proved to be better than those of Amancio (2013) and Silveira (2014), which in the validation phase, They showed correlations of 0.89 and 0.94 respectively. The final model of this work has an architecture comprised of 10 nodes in the input layer, 34 neurons distributed throughout four hidden layers, and one neuron in the output layer (A: 10-15-9-7-3-1) using to calculate the average number of NSPT representative values along the cutting shaft.
39
Campos, Lucio Flavio de Albuquerque. "Classificação de Lesões em Mamografias Digitais Utilizando Análise de Componentes Independentes e Perceptron Multicamadas." Universidade Federal do Maranhão, 2006. http://tedebc.ufma.br:8080/jspui/handle/tede/344.
Full textAbstract:
Made available in DSpace on 2016-08-17T14:52:50Z (GMT). No. of bitstreams: 1
Lucio Flavio de Albuquerque Campos.pdf: 558329 bytes, checksum: 1be43a8cd03ed147f8b5187d7e538e5a (MD5)
Previous issue date: 2006-03-24We propose a method for discrimination and classification of mammograms with benign, malignant and normal tissues using independent component analysis and neural networks. The method was tested for a mammogram set from MIAS database, and multilayer perceptron. The method obtained a success rate of 97.83% , with 97.5% of specificity and 98% of sensitivity.
Neste trabalho, propomos um método para discriminação e classificação de mamogramas, com diagnóstico maligno, benigno e normal, usando análise de componentes independentes e redes neurais. O método foi testado com mamogramas da MIAS database, e com redes perceptron multicamadas. O método obteve uma taxa de sucesso média de 97.83%, com 97.5% de especificidade, e 98% de sensibilidade.
40
Araújo, Carla Beatriz Costa de. "Aplicação das redes neurais artificiais do tipo perceptron na estimativa de recalques em estacas." reponame:Repositório Institucional da UFC, 2015. http://www.repositorio.ufc.br/handle/riufc/12932.
Full textAbstract:
ARAÚJO, C. B. C. Aplicação das redes neurais artificiais do tipo perceptron na estimativa de recalques em estacas. 2015. 203 f. Dissertação (Mestrado em Engenharia Civil: Geotecnia) – Centro de Tecnologia, Universidade Federal do Ceará, Fortaleza, 2015.Submitted by Marlene Sousa (mmarlene@ufc.br) on 2015-06-24T13:04:33Z No. of bitstreams: 1 2015_dis_cbcaraújo.pdf: 11710979 bytes, checksum: e4ca4d2bd1394da9c96e33c49a57ee7c (MD5)
Approved for entry into archive by Marlene Sousa(mmarlene@ufc.br) on 2015-06-24T19:02:43Z (GMT) No. of bitstreams: 1 2015_dis_cbcaraújo.pdf: 11710979 bytes, checksum: e4ca4d2bd1394da9c96e33c49a57ee7c (MD5)
Made available in DSpace on 2015-06-24T19:02:43Z (GMT). No. of bitstreams: 1 2015_dis_cbcaraújo.pdf: 11710979 bytes, checksum: e4ca4d2bd1394da9c96e33c49a57ee7c (MD5) Previous issue date: 2015-04-24
Use of artificial neural networks (ANN) in the estimation of settlements in foundations deep has proven an effective tool. The work of Amancio (2013) and Silveira (2014), the use of RNA showed good results for predicting settlements in continuous stakes propellers, metal piles driven and bored piles. However, some modeled stakes had far behavior of real results, where modeling results indicate sharp increases in stiffness soil-cutting system. In this research, it developed a model with a neural network of the multilayer perceptron to improve the performance of the models Amâncio (2013) and Silveira (2014). To development work initially polls results of analyzes were made Percussion SPT and static load tests of 199 stakes used at work presented by Silveira (2014), making up an assessment of the consistency of the information, in order to have a more heterogeneous and the representative assembly. After conducting changes, has come up with a set with 141 stakes, totaling 1,320 examples of the type entrance exit. Were defined as model input variables: the type of pile, the length of the pile, the pile diameter, the number of representative values when NSPT Over stake stem (called NF), the NSPT on the edge of the pile, depth of the layer the influence of load relative to the cutting edge, the factor representative of the soil layers clay, the representative factor of silty soil layers, the representative factor of the layers sandy soil and the applied load. Four different ways of calculation have been studied in NF input variable, which are: sum, average, weighted sum and weighted average. With input variables presented were worked models where the output variable was the repression of deep foundation. The modeling of RNA was made using the QNET program 2000 and were carried out training and validation of different architectures. The model had better performance showed correlation coefficient between the actual settlements and settlements modeled in the training of 0.99 and 0.98 in the validation. The results proved to be better than those of Amancio (2013) and Silveira (2014), which in the validation phase, They showed correlations of 0.89 and 0.94 respectively. The final model of this work has an architecture comprised of 10 nodes in the input layer, 34 neurons distributed throughout four hidden layers, and one neuron in the output layer (A: 10-15-9-7-3-1) using to calculate the average number of NSPT representative values along the cutting shaft
A utilização das redes neurais artificiais (RNA) na estimativa de recalques em fundações profundas é comprovadamente uma ferramenta eficiente. Nos trabalhos de Amâncio (2013) e Silveira (2014), o emprego das RNA apresentou bons resultados para a previsão de recalques em estacas hélices contínuas, estacas cravadas metálicas e estacas escavadas. Porém, algumas estacas modeladas apresentaram comportamento muito distante dos resultados reais, onde os resultados da modelagem indicaram aumentos bruscos na rigidez do sistema solo-estaca. Nesta pesquisa, foi desenvolvido um modelo com uma rede neural do tipo perceptron multicamadas de forma a melhorar o desempenho dos modelos de Amâncio (2013) e Silveira (2014). Para desenvolvimento do trabalho, inicialmente foram feitas análises dos resultados de sondagens à percussão do tipo SPT e provas de carga estáticas das 199 estacas utilizadas no trabalho apresentado por Silveira (2014), fazendo-se uma avaliação da consistência das informações, com o objetivo de ter um conjunto mais heterogêneo e representativo. Após a realização de alterações, chegou-se a um conjunto com 141 estacas, totalizando 1.320 exemplos do tipo entrada-saída. Foram definidas como variáveis de entrada do modelo: o tipo de estaca, o comprimento da estaca, o diâmetro da estaca, o número representativo dos valores de NSPT ao longo do fuste da estaca (denominada NF), o NSPT na ponta da estaca, profundidade da camada de influência da carga em relação a ponta da estaca, o fator representativo das camadas de solo argiloso, o fator representativo das camadas de solo siltoso, o fator representativo das camadas de solo arenoso e a carga aplicada. Foram estudadas quatro diferentes formas de cálculo da variável de entrada NF, sendo estas: soma, média, soma ponderada e média ponderada. Com as variáveis de entrada apresentadas foram trabalhados modelos onde a variável de saída fosse o recalque da fundação profunda. A modelagem das RNA foi feita utilizando o programa QNET 2000, e foram realizados o treinamento e a validação de diferentes arquiteturas. O modelo que teve melhor desempenho apresentou coeficiente de correlação entre os recalques reais e os recalques modelados no treinamento de 0,99 e na validação de 0,98. Os resultados obtidos mostraram-se melhores que os de Amâncio (2013) e Silveira (2014), que na fase de validação, apresentaram correlações de 0,89 e 0,94 respectivamente. O modelo final deste trabalho possui uma arquitetura formada por 10 nós na camada de entrada, 34 neurônios distribuídos ao longo de quatro camadas ocultas e um neurônio na camada de saída (A:10-15-9-7-3-1), utilizando a média para cálculo do número representativo dos valores de NSPT ao longo do fuste da estaca
41
Diouf, Daouda. "Méthode mixte d'inversion neuro-variationnelle d'images de la couleur de l'océan : Application aux signaux SeaWIFS au large de l'Afrique de l'Ouest." Paris 6, 2012. http://www.theses.fr/2012PA066181.
Full textAbstract:
Les capteurs optiques, destinés à observer l’océan depuis l’espace, mesurent le rayonnement solaire réfléchi vers l’espace par le système océan-atmosphère. La réflectance marine intéressante pour l’analyse de l’océan représente en moyenne au plus 10% de la lumière totale reçue par le capteur et est obtenue à l’issue du processus de correction atmosphérique. L’inversion de ce signal marin permet d’obtenir les paramètres géophysiques intéressants pour l’étude de l’océan, tels que la concentration en chlorophylles-a, pigment principal du phytoplancton. En général la difficulté des algorithmes standards de correction atmosphériques réside dans la quantification de l’impact des aérosols présents dans l’atmosphère sur le signal mesuré par le capteur surtout lorsqu’ils sont absorbants. Nous proposons des méthodologies statistico-mathématiques adaptés afin de déterminer les types d’aérosols atmosphériques et leurs épaisseurs optiques et ensuite restituer la couleur de l’océan. Cette méthodologie qui est une combinaison de plusieurs algorithmes neuronaux et d’une optimisation variationnelle est appelée SOM-NV et a été appliquée sur treize années d’observations du capteur SeaWiFS au large de l’Afrique de l’Ouest. Les épaisseurs optiques et les coefficients d’Angström mesurés in-situ (mesures AERONET) ont permis de valider respectivement les épaisseurs optiques et les types d’aérosols obtenues par SOM-NV. D’autre part la méthode est aussi capable de détecter les aérosols absorbants tels que les poussières sahariennes et donne des résultats précis pour les valeurs d'épaisseur optiques supérieures à 0,35, ce qui n'est pas le cas pour le produit standard SeaWiFSOptical sensors, used to observe the ocean from space, measure the solar radiation reflected back to space by the ocean-atmosphere system. The marine reflectance interesting for the analysis of the ocean represents an average at most 10% of the total light received by the sensor and is obtained at the end of an atmospheric correction process. The inversion of this marine signal provides geophysical parameters interesting for the study of the ocean, such as the chlorophyll-a concentration, a major pigment of phytoplankton. In general the difficulty of standard atmospheric correction algorithms lies in quantifying the impact of aerosols in the atmosphere on the signal measured by the sensor especially when they are absorbing. We present adapted statistical and mathematical methodologies to determine atmospheric aerosols types and their optical thickness and then retrieve the ocean color. This methodology which is a combination of several neural network algorithms and a variational optimization is called SOM-NV and was applied to thirteen years of SeaWiFS observations off West Africa. The aerosols optical thickness and Angström coefficients measured in-situ (AERONET measurements) were respectively used to validate the aerosols optical thickness and aerosols types obtained by SOM-NV. Furthermore the method is also able to detect absorbing aerosols such as Saharan dust and gives accurate results for optical thickness values greater than 0. 35, which is not the case for SeaWiFS standard product
42
Shaheed, Mohammad Hasan. "Neural and genetic modelling, control and real-time finite simulation of flexible manipulators." Thesis, University of Sheffield, 2000. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.327649.
Full text
43
Esteves, João Trevizoli. "Climate and agrometeorology forecasting using soft computing techniques. /." Jaboticabal, 2018. http://hdl.handle.net/11449/180833.
Full textAbstract:
Orientador: Glauco de Souza RolimResumo: Precipitação, em pequenas escalas de tempo, é um fenômeno associado a altos níveis de incerteza e variabilidade. Dada a sua natureza, técnicas tradicionais de previsão são dispendiosas e exigentes em termos computacionais. Este trabalho apresenta um modelo para prever a ocorrência de chuvas em curtos intervalos de tempo por Redes Neurais Artificiais (RNAs) em períodos acumulados de 3 a 7 dias para cada estação climática, mitigando a necessidade de predizer o seu volume. Com essa premissa pretende-se reduzir a variância, aumentar a tendência dos dados diminuindo a responsabilidade do algoritmo que atua como um filtro para modelos quantitativos, removendo ocorrências subsequentes de valores de zero(ausência) de precipitação, o que influencia e reduz seu desempenho. O modelo foi desenvolvido com séries temporais de 10 regiões agricolamente relevantes no Brasil, esses locais são os que apresentam as séries temporais mais longas disponíveis e são mais deficientes em previsões climáticas precisas, com 60 anos de temperatura média diária do ar e precipitação acumulada. foram utilizados para estimar a evapotranspiração potencial e o balanço hídrico; estas foram as variáveis utilizadas como entrada para as RNAs. A precisão média para todos os períodos acumulados foi de 78% no verão, 71% no inverno 62% na primavera e 56% no outono, foi identificado que o efeito da continentalidade, o efeito da altitude e o volume da precipitação normal , tem um impacto direto na precisão das RNAs. Os... (Resumo completo, clicar acesso eletrônico abaixo)
Abstract: Precipitation, in short periods of time, is a phenomenon associated with high levels of uncertainty and variability. Given its nature, traditional forecasting techniques are expensive and computationally demanding. This paper presents a model to forecast the occurrence of rainfall in short ranges of time by Artificial Neural Networks(ANNs) in accumulated periods from 3 to 7 days for each climatic season, mitigating the necessity of predicting its amount. With this premise it is intended to reduce the variance, rise the bias of data and lower the responsibility of the model acting as a filter for quantitative models by removing subsequent occurrences of zeros values of rainfall which leads to bias the and reduces its performance. The model were developed with time series from 10 agriculturally relevant regions in Brazil, these places are the ones with the longest available weather time series and and more deficient in accurate climate predictions, it was available 60 years of daily mean air temperature and accumulated precipitation which were used to estimate the potential evapotranspiration and water balance; these were the variables used as inputs for the ANNs models. The mean accuracy of the model for all the accumulated periods were 78% on summer, 71% on winter 62% on spring and 56% on autumn, it was identified that the effect of continentality, the effect of altitude and the volume of normal precipitation, have a direct impact on the accuracy of the ANNs. The models have ... (Complete abstract click electronic access below)
Mestre
44
Ashok, Ashish Kumar. "Predictive data mining in a collaborative editing system: the Wikipedia articles for deletion process." Thesis, Kansas State University, 2011. http://hdl.handle.net/2097/12026.
Full textAbstract:
Master of ScienceDepartment of Computing and Information Sciences
William H. Hsu
In this thesis, I examine the Articles for Deletion (AfD) system in /Wikipedia/, a large-scale collaborative editing project. Articles in Wikipedia can be nominated for deletion by registered users, who are expected to cite criteria for deletion from the Wikipedia deletion. For example, an article can be nominated for deletion if there are any copyright violations, vandalism, advertising or other spam without relevant content, advertising or other spam without relevant content. Articles whose subject matter does not meet the notability criteria or any other content not suitable for an encyclopedia are also subject to deletion. The AfD page for an article is where Wikipedians (users of Wikipedia) discuss whether an article should be deleted. Articles listed are normally discussed for at least seven days, after which the deletion process proceeds based on community consensus. Then the page may be kept, merged or redirected, transwikied (i.e., copied to another Wikimedia project), renamed/moved to another title, userfied or migrated to a user subpage, or deleted per the deletion policy. Users can vote to keep, delete or merge the nominated article. These votes can be viewed in article’s view AfD page. However, this polling does not necessarily determine the outcome of the AfD process; in fact, Wikipedia policy specifically stipulates that a vote tally alone should not be considered sufficient basis for a decision to delete or retain a page. In this research, I apply machine learning methods to determine how the final outcome of an AfD process is affected by factors such as the difference between versions of an article, number of edits, and number of disjoint edits (according to some contiguity constraints). My goal is to predict the outcome of an AfD by analyzing the AfD page and editing history of the article. The technical objectives are to extract features from the AfD discussion and version history, as reflected in the edit history page, that reflect factors such as those discussed above, can be tested for relevance, and provide a basis for inductive generalization over past AfDs. Applications of such feature analysis include prediction and recommendation, with the performance goal of improving the precision and recall of AfD outcome prediction.
45
Vural, Hulya. "Comparison Of Rough Multi Layer Perceptron And Rough Radial Basis Function Networks Using Fuzzy Attributes." Master's thesis, METU, 2004. http://etd.lib.metu.edu.tr/upload/12605293/index.pdf.
Full textAbstract:
The hybridization of soft computing methods of Radial Basis Function (RBF) neural networks, Multi Layer Perceptron (MLP) neural networks with back-propagation learning, fuzzy sets and rough sets are studied in the scope of this thesis. Conventional MLP, conventional RBF, fuzzy MLP, fuzzy RBF, rough fuzzy MLP, and rough fuzzy RBF networks are compared. In the fuzzy neural networks implemented in this thesis, the input data and the desired outputs are given fuzzy membership values as the fuzzy properties &ldquolow&rdquo
, &ldquo
medium&rdquo
and &ldquo
high&rdquo
. In the rough fuzzy MLP, initial weights and near optimal number of hidden nodes are estimated using rough dependency rules. A rough fuzzy RBF structure similar to the rough fuzzy MLP is proposed. The rough fuzzy RBF was inspected whether dependencies like the ones in rough fuzzy MLP can be concluded.
46
Evans, John Thomas. "Investigation of a multi-layer perceptron network to model and control a non-linear system." Thesis, Liverpool John Moores University, 1994. http://researchonline.ljmu.ac.uk/4945/.
Full textAbstract:
This thesis describes the development and implementation of an on-line optimal predictive controller incorporating a neural network model of a non-linear process. The scheme is based on a Multi-Layer Perceptron neural net-work as a modelling tool for a real non-linear, dual tank, liquid level process. A neural network process model is developed and evaluated firstly in simulation studies and then subsequently on the real process. During the development of the network model, the ability of the network to predict the process output multiple time steps ahead was investigated. This led to investigations into a number of important aspects such as the network topology, training algorithms, period of network training, model validation and conditioning of the process data. Once the development of the neural network model had been achieved, it was included into a predictive control scheme where an on-line comparison with a conventional three term controller was undertaken. Improvements in process control performance that can be achieved in practice using a neural control scheme are illustrated. Additionally, an insight into the dynamics and stability of the neural control scheme was obtained in a novel application of linear system identification techniques. The research shows that a technique of conditioning the process data, called spread encoding, enabled a neural network to accurately emulate the real process using only process input information and this facilitated accurate multi-step-ahead predictive control to be performed.
47
Gao, Zhenning. "Parallel and Distributed Implementation of A Multilayer Perceptron Neural Network on A Wireless Sensor Network." University of Toledo / OhioLINK, 2014. http://rave.ohiolink.edu/etdc/view?acc_num=toledo1383764269.
Full text
48
Dlugosz, Stephan. "Multi-layer perceptron networks for ordinal data analysis : order independent online learning by sequential estimation /." Berlin : Logos, 2008. http://d-nb.info/990567311/04.
Full text
49
Santos, Diego Santiago dos. "Utilização da tecnologia bluetooth associada a redes neurais artificiais (PMC) para monitoramento e rastreamento de suínos." Universidade de São Paulo, 2014. http://www.teses.usp.br/teses/disponiveis/74/74131/tde-22092014-135707/.
Full textAbstract:
O presente trabalho teve como objetivo apresentar uma metodologia que permita encontrar o posicionamento e rastrear as diferentes localizações de um suíno em uma baia, utilizando o valor do Receiver Signal Strenght Indicator (RSSI), entre o dispositivo móvel (suíno) e três dispositivos fixos, e uma Rede Neural Artificial do tipo Perceptron Multicamadas (PMC), responsável por interpretar os sinais RSSI e transformá-los em valores conhecidos, como em um plano cartesiano, com coordenadas no eixo X e eixo Y. A região de teste foi dividida em 289 pontos, sendo 286 utilizados para coleta de dados e para o treinamento da rede PMC. Para cada ponto, foram armazenados a sua posição dentro da baia e o valor RSSI entre o dispositivo móvel e os três dispositivos fixos. O processo foi repetido para 8 pontos escolhidos aleatoriamente dentro do espaço de teste e inseridos como entradas na rede PMC. Após treinamentos e operações realizadas com diversas arquiteturas foi possível concluir que àquela dotada de 10 neurônios na camada intermediária consistiu na melhor alternativa, cujos resultados de monitoramento e rastreamento das posições do dispositivo móvel foram encontradas com valores aceitáveis de exatidão.This paper aims to present a methodology to find the positioning and tracking of the different locations of a pig in a stall, using the value of the Receiver Signal Strength Indicator (RSSI), between the mobile device (pig) and three devices fixed, and an Artificial Neural Network Multilayer Perceptron (MLP), responsible for interpreting the RSSI signals and turning them into known values, such as on a Cartesian plane, with coordinates on X axis and Y axis. The test region was divided into 289 points, with 286 points used for data collection and training of PMC network, and for each point, it was stored its position inside the stall and its RSSI value between the mobile devices and the three fixed. The process was repeated for 8 points chosen randomly within the space of test and entered as inputs into the PMC network. After training and operations with various architectures it was concluded that the architecture with 10 neurons in the hidden layer was the best alternative, whose the results of monitoring and tracking the positions of mobile device were found with acceptable accuracy.
50
Gaspar, Thiago Lombardi. "Reconhecimento de faces humanas usando redes neurais MLP." Universidade de São Paulo, 2006. http://www.teses.usp.br/teses/disponiveis/18/18133/tde-27042006-231620/.
Full textAbstract:
O objetivo deste trabalho foi desenvolver um algoritmo baseado em redes neurais para o reconhecimento facial. O algoritmo contém dois módulos principais, um módulo para a extração de características e um módulo para o reconhecimento facial, sendo aplicado sobre imagens digitais nas quais a face foi previamente detectada. O método utilizado para a extração de características baseia-se na aplicação de assinaturas horizontais e verticais para localizar os componentes faciais (olhos e nariz) e definir a posição desses componentes. Como entrada foram utilizadas imagens faciais de três bancos distintos: PICS, ESSEX e AT&T. Para esse módulo, a média de acerto foi de 86.6%, para os três bancos de dados. No módulo de reconhecimento foi utilizada a arquitetura perceptron multicamadas (MLP), e para o treinamento dessa rede foi utilizado o algoritmo de aprendizagem backpropagation. As características faciais extraídas foram aplicadas nas entradas dessa rede neural, que realizou o reconhecimento da face. A rede conseguiu reconhecer 97% das imagens que foram identificadas como pertencendo ao banco de dados utilizado. Apesar dos resultados satisfatórios obtidos, constatou-se que essa rede não consegue separar adequadamente características faciais com valores muito próximos, e portanto, não é a rede mais eficiente para o reconhecimento facialThis research presents a facial recognition algorithm based in neural networks. The algorithm contains two main modules: one for feature extraction and another for face recognition. It was applied in digital images from three database, PICS, ESSEX and AT&T, where the face was previously detected. The method for feature extraction was based on previously knowledge of the facial components location (eyes and nose) and on the application of the horizontal and vertical signature for the identification of these components. The mean result obtained for this module was 86.6% for the three database. For the recognition module it was used the multilayer perceptron architecture (MLP), and for training this network it was used the backpropagation algorithm. The extracted facial features were applied to the input of the neural network, that identified the face as belonging or not to the database with 97% of hit ratio. Despite the good results obtained it was verified that the MLP could not distinguish facial features with very close values. Therefore the MLP is not the most efficient network for this task