Atualmente diversos sistemas sensoriais fornecem dados em fluxos e essas observações medidas são frequentemente de alta dimensionalidade, ou seja, o número de variáveis medidas é grande, e as observações chegam em sequência. Este é, em particular, o caso de sistemas de visão em robôs. Aprendizagem supervisionada e não-supervisionada com esses fluxos de dados é um desafio, porque o algoritmo deve ser capaz de aprender com cada observação e depois descartá-la antes de considerar a próxima, mas diversos métodos requerem todo o conjunto de dados a fim de estimar seus parâmetros e, portanto, não são adequados para aprendizagem em tempo real. Além disso, muitas abordagens sofrem com a denominada maldição da dimensionalidade (BELLMAN, 1961) e não conseguem lidar com dados de entrada de alta dimensionalidade. Para superar os problemas descritos anteriormente, este trabalho propõe um novo modelo de rede neural probabilístico e incremental, denominado Local Projection Incremental Gaussian Mixture Network (LP-IGMN), que é capaz de realizar aprendizagem perpétua com dados de alta dimensionalidade, ou seja, ele pode aprender continuamente considerando a estabilidade dos parâmetros do modelo atual e automaticamente ajustar sua topologia levando em conta a fronteira do subespaço encontrado por cada neurônio oculto. O método proposto pode encontrar o subespaço intrísico onde os dados se localizam, o qual é denominado de subespaço principal. Ortogonal ao subespaço principal, existem as dimensões que são ruidosas ou que carregam pouca informação, ou seja, com pouca variância, e elas são descritas por um único parâmetro estimado. Portanto, LP-IGMN é robusta a diferentes fontes de dados e pode lidar com grande número de variáveis ruidosas e/ou irrelevantes nos dados medidos. Para avaliar a LP-IGMN nós realizamos diversos experimentos usando conjunto de dados simulados e reais. Demonstramos ainda diversas aplicações do nosso método em tarefas de reconhecimento de imagens. Os resultados mostraram que o desempenho da LP-IGMN é competitivo, e geralmente superior, com outras abordagens do estado da arte, e que ela pode ser utilizada com sucesso em aplicações que requerem aprendizagem perpétua em espaços de alta dimensionalidade. / Nowadays several sensory systems provide data in ows and these measured observations are frequently high-dimensional, i.e., the number of measured variables is large, and the observations are arriving in a sequence. This is in particular the case of robot vision systems. Unsupervised and supervised learning with such data streams is challenging, because the algorithm should be capable of learning from each observation and then discard it before considering the next one, but several methods require the whole dataset in order to estimate their parameters and, therefore, are not suitable for online learning. Furthermore, many approaches su er with the so called curse of dimensionality (BELLMAN, 1961) and can not handle high-dimensional input data. To overcome the problems described above, this work proposes a new probabilistic and incremental neural network model, called Local Projection Incremental Gaussian Mixture Network (LP-IGMN), which is capable to perform life-long learning with high-dimensional data, i.e., it can continuously learn considering the stability of the current model's parameters and automatically adjust its topology taking into account the subspace's boundary found by each hidden neuron. The proposed method can nd the intrinsic subspace where the data lie, which is called the principal subspace. Orthogonal to the principal subspace, there are the dimensions that are noisy or carry little information, i.e., with small variance, and they are described by a single estimated parameter. Therefore, LP-IGMN is robust to di erent sources of data and can deal with large number of noise and/or irrelevant variables in the measured data. To evaluate LP-IGMN we conducted several experiments using simulated and real datasets. We also demonstrated several applications of our method in image recognition tasks. The results have shown that the LP-IGMN performance is competitive, and usually superior, with other stateof- the-art approaches, and it can be successfully used in applications that require life-long learning in high-dimensional spaces.

Atualmente diversos sistemas sensoriais fornecem dados em fluxos e essas observações medidas são frequentemente de alta dimensionalidade, ou seja, o número de variáveis medidas é grande, e as observações chegam em sequência. Este é, em particular, o caso de sistemas de visão em robôs. Aprendizagem supervisionada e não-supervisionada com esses fluxos de dados é um desafio, porque o algoritmo deve ser capaz de aprender com cada observação e depois descartá-la antes de considerar a próxima, mas diversos métodos requerem todo o conjunto de dados a fim de estimar seus parâmetros e, portanto, não são adequados para aprendizagem em tempo real. Além disso, muitas abordagens sofrem com a denominada maldição da dimensionalidade (BELLMAN, 1961) e não conseguem lidar com dados de entrada de alta dimensionalidade. Para superar os problemas descritos anteriormente, este trabalho propõe um novo modelo de rede neural probabilístico e incremental, denominado Local Projection Incremental Gaussian Mixture Network (LP-IGMN), que é capaz de realizar aprendizagem perpétua com dados de alta dimensionalidade, ou seja, ele pode aprender continuamente considerando a estabilidade dos parâmetros do modelo atual e automaticamente ajustar sua topologia levando em conta a fronteira do subespaço encontrado por cada neurônio oculto. O método proposto pode encontrar o subespaço intrísico onde os dados se localizam, o qual é denominado de subespaço principal. Ortogonal ao subespaço principal, existem as dimensões que são ruidosas ou que carregam pouca informação, ou seja, com pouca variância, e elas são descritas por um único parâmetro estimado. Portanto, LP-IGMN é robusta a diferentes fontes de dados e pode lidar com grande número de variáveis ruidosas e/ou irrelevantes nos dados medidos. Para avaliar a LP-IGMN nós realizamos diversos experimentos usando conjunto de dados simulados e reais. Demonstramos ainda diversas aplicações do nosso método em tarefas de reconhecimento de imagens. Os resultados mostraram que o desempenho da LP-IGMN é competitivo, e geralmente superior, com outras abordagens do estado da arte, e que ela pode ser utilizada com sucesso em aplicações que requerem aprendizagem perpétua em espaços de alta dimensionalidade. / Nowadays several sensory systems provide data in ows and these measured observations are frequently high-dimensional, i.e., the number of measured variables is large, and the observations are arriving in a sequence. This is in particular the case of robot vision systems. Unsupervised and supervised learning with such data streams is challenging, because the algorithm should be capable of learning from each observation and then discard it before considering the next one, but several methods require the whole dataset in order to estimate their parameters and, therefore, are not suitable for online learning. Furthermore, many approaches su er with the so called curse of dimensionality (BELLMAN, 1961) and can not handle high-dimensional input data. To overcome the problems described above, this work proposes a new probabilistic and incremental neural network model, called Local Projection Incremental Gaussian Mixture Network (LP-IGMN), which is capable to perform life-long learning with high-dimensional data, i.e., it can continuously learn considering the stability of the current model's parameters and automatically adjust its topology taking into account the subspace's boundary found by each hidden neuron. The proposed method can nd the intrinsic subspace where the data lie, which is called the principal subspace. Orthogonal to the principal subspace, there are the dimensions that are noisy or carry little information, i.e., with small variance, and they are described by a single estimated parameter. Therefore, LP-IGMN is robust to di erent sources of data and can deal with large number of noise and/or irrelevant variables in the measured data. To evaluate LP-IGMN we conducted several experiments using simulated and real datasets. We also demonstrated several applications of our method in image recognition tasks. The results have shown that the LP-IGMN performance is competitive, and usually superior, with other stateof- the-art approaches, and it can be successfully used in applications that require life-long learning in high-dimensional spaces.

Submitted by Glauco Gomes de Azevedo (glaucogazevedo@gmail.com) on 2018-08-28T20:54:50Z No. of bitstreams: 1 dissertacao_glauco_azevedo.pdf: 1909706 bytes, checksum: 6636e75aa9da1db2615932f064fd1138 (MD5) / Approved for entry into archive by Janete de Oliveira Feitosa (janete.feitosa@fgv.br) on 2018-09-10T19:38:08Z (GMT) No. of bitstreams: 1 dissertacao_glauco_azevedo.pdf: 1909706 bytes, checksum: 6636e75aa9da1db2615932f064fd1138 (MD5) / Made available in DSpace on 2018-09-12T17:39:51Z (GMT). No. of bitstreams: 1 dissertacao_glauco_azevedo.pdf: 1909706 bytes, checksum: 6636e75aa9da1db2615932f064fd1138 (MD5) Previous issue date: 2018-06-04 / Neste trabalho é proposta uma metodologia para estimar distâncias entre pontos de dados mistos, contínuos e categóricos, contendo dados faltantes. Estimação de distâncias é a base para muitos métodos de regressão/classificação, tais como vizinhos mais próximos e análise de discriminantes, e para técnicas de clusterização como k-means e k-medoids. Métodos clássicos para manipulação de dados faltantes se baseiam em imputação pela média, o que pode subestimar a variância, ou em métodos baseados em regressão. Infelizmente, quando a meta é a estimar a distância entre observações, a imputação de dados pode performar de modo ineficiente e enviesar os resultados na direção do modelo. Na proposta desse trabalho, estima-se a distância dos pares diretamente, tratando os dados faltantes como aleatórios. A distribuição conjunta dos dados é aproximada utilizando um modelo de mistura multivariado para dados mistos, contínuos e categóricos. Apresentamentos um algoritmo do tipo EM para estimar a mistura e uma metodologia geral para estimar a distância entre observações. Simulações mostram que um método proposto performa tanto dados simulados, como reais. / In this work we propose a methodology to estimate the pairwise distance between mixed continuous and categorical data with missing values. Distance estimation is the base for many regression/classification methods, such as nearest neighbors and discriminant analysis, and for clustering techniques such as k-means and k-medoids. Classical methods for handling missing data rely on mean imputation, that could underestimate the variance, or regression-based imputation methods. Unfortunately, when the goal is to estimate the distance between observations, data imputation may perform badly and bias the results toward the data imputation model. In this work we estimate the pairwise distances directly, treating the missing data as random. The joint distribution of the data is approximated using a multivariate mixture model for mixed continuous and categorical data. We present an EM-type algorithm for estimating the mixture and a general methodology for estimating the distance between observations. Simulation shows that the proposed method performs well in both simulated and real data.

A common problem when using background models to segment moving objects from video sequences is that objects cast shadow usually significantly differ from the background and therefore get detected as foreground. This causes several problems when extracting and labeling objects, such as object shape distortion and several objects merging together. The purpose of this thesis is to explore various possibilities to handle this problem. Three methods for statistical background modeling are reviewed. All methods work on a per pixel basis, the first is based on approximating the median, the next on using Gaussian mixture models, and the last one is based on channel representation. It is concluded that all methods detect cast shadows as foreground. A study of existing methods to handle cast shadows has been carried out in order to gain knowledge on the subject and get ideas. A common approach is to transform the RGB-color representation into a representation that separates color into intensity and chromatic components in order to determine whether or not newly sampled pixel-values are related to the background. The color spaces HSV, IHSL, CIELAB, YCbCr, and a color model proposed in the literature (Horprasert et al.) are discussed and compared for the purpose of shadow detection. It is concluded that Horprasert's color model is the most suitable for this purpose. The thesis ends with a proposal of a method to combine background modeling using Gaussian mixture models with shadow detection using Horprasert's color model. It is concluded that, while not perfect, such a combination can be very helpful in segmenting objects and detecting their cast shadow.

Les bases de données modernes sont de très grande taille, parfois divisées et distribuées sur plusieurs lieux de stockage, ou encore sous forme de flux de données : ceci soulève de nouveaux défis majeurs pour les méthodes d’apprentissage statistique. Une des méthodes récentes capable de s’adapter à ces situations consiste à d’abord compresser les données en une structure appelée sketch linéaire, puis ensuite de réaliser la tâche d’apprentissage en utilisant uniquement ce sketch, ce qui est extrêmement rapide si celui-ci est de petite taille. Dans cette thèse, nous définissons une telle méthode pour estimer un modèle de mélange de distributions de probabilités à partir des données, en utilisant uniquement un sketch de celles-ci. Ce sketch est défini en s’inspirant de plusieurs notions venant du domaine des méthodes à noyaux : le plongement par noyau moyen et les approximations aléatoires de noyaux. Défini comme tel, le sketch correspond à des mesures linéaires de la distribution de probabilité sous-jacente aux données. Ainsi nous analysons le problème en utilisant des outils venant du domaine de l’acquisition comprimée, dans lequel un signal est mesuré aléatoirement sans perte d’information, sous certaines conditions. Nous étendons certains résultats de l’acquisition comprimée à la dimension infinie, donnons des conditions génériques garantissant le succès de notre méthode d’estimation de modèles de mélanges, et les appliquons à plusieurs problèmes, dont notamment celui d’estimer des mélanges de distributions stables multivariées, pour lequel il n’existait à ce jour aucun estimateur. Notre analyse est basée sur la construction d’opérateurs de sketch construits aléatoirement, qui satisfont une Propriété d’Isométrie Restreinte dans l’espace de Banach des mesures finies signées avec forte probabilité. Dans une second partie, nous introduisons un algorithme glouton capable heuristiquement d’estimer un modèle de mélange depuis un sketch linéaire. Cet algorithme est appliqué sur données simulées et réelles à trois problèmes : l’estimation de centres significatifs dans les données, pour lequel on constate que la méthode de sketch est significativement plus rapide qu’un algorithme de k-moyennes classique, l’estimation de mélanges de Gaussiennes, pour lequel elle est plus rapide qu’un algorithme d’Espérance-Maximisation, et enfin l’estimation de mélange de distributions stables multivariées, pour lequel il n’existait à ce jour, à notre connaissance, aucun algorithme capable de réaliser une telle tâche. / Learning parameters from voluminous data can be prohibitive in terms of memory and computational requirements. Furthermore, new challenges arise from modern database architectures, such as the requirements for learning methods to be amenable to streaming, parallel and distributed computing. In this context, an increasingly popular approach is to first compress the database into a representation called a linear sketch, that satisfies all the mentioned requirements, then learn the desired information using only this sketch, which can be significantly faster than using the full data if the sketch is small. In this thesis, we introduce a generic methodology to fit a mixture of probability distributions on the data, using only a sketch of the database. The sketch is defined by combining two notions from the reproducing kernel literature, namely kernel mean embedding and Random Features expansions. It is seen to correspond to linear measurements of the underlying probability distribution of the data, and the estimation problem is thus analyzed under the lens of Compressive Sensing (CS), in which a (traditionally finite-dimensional) signal is randomly measured and recovered. We extend CS results to our infinite-dimensional framework, give generic conditions for successful estimation and apply them analysis to many problems, with a focus on mixture models estimation. We base our method on the construction of random sketching operators such that some Restricted Isometry Property (RIP) condition holds in the Banach space of finite signed measures with high probability. In a second part we introduce a flexible heuristic greedy algorithm to estimate mixture models from a sketch. We apply it on synthetic and real data on three problems: the estimation of centroids from a sketch, for which it is seen to be significantly faster than k-means, Gaussian Mixture Model estimation, for which it is more efficient than Expectation-Maximization, and the estimation of mixtures of multivariate stable distributions, for which, to our knowledge, it is the only algorithm capable of performing such a task.

Mixture experiments are widely applied. The Scheffé quadratic polynomial is the most popular mixture model in industry due to its simplicity, but it fails to accurately describe the behaviour of response variables that deviate greatly from linear blending. Higherorder Scheffé polynomials do possess the ability to predict such behaviour but become increasingly more complex to use and the number of estimable parameters grow exponentially [15]. A parameter-parsimonious mixture model, developed from the linear blending rule with weighted power means and Wohl's Q-fractions, is introduced. Bootstrap is employed to analyse the model statistically. The model is proved to be flexible enough to model non-linear deviations from linear blending without losing the simplicity of the linear blending rule. / Dissertation (MSc)--University of Pretoria, 2011. / Chemical Engineering / unrestricted

Orientador: Fernando José Von Zuben / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação / Made available in DSpace on 2018-08-18T10:35:24Z (GMT). No. of bitstreams: 1 Goncalves_AndreRicardo_M.pdf: 2010410 bytes, checksum: d89f7061364f054a7b11d52cc61d27a4 (MD5) Previous issue date: 2011 / Resumo: O dinamismo do mundo moderno traz consigo grandes desafios científicos e tecnológicos, particularmente junto a problemas de otimização. Problemas antes tratados de forma estática estão sendo reformulados para incorporar esse dinamismo, exigindo com isso novas estratégias de solução. Meta-heurísticas populacionais para otimização surgem então como abordagens promissoras, visto que favorecem a exploração do espaço de busca e contribuem para a adaptação ao dinamismo do ambiente. Foram tratados aqui algoritmos de estimação de distribuição (AEDs), os quais empregam modelos probabilísticos para identificar regiões promissoras do espaço de busca. Pelo fato de serem raras e limitadas as propostas de AEDs para problemas dinâmicos, principalmente em espaços de busca contínuos, foram concebidos AEDs baseados em modelos de mistura gaussianos flexíveis, auto-controláveis e com baixo custo computacional, incluindo ainda operadores de manutenção de diversidade e de controle de convergência. Uma extensa comparação com métodos alternativos de otimização para ambientes dinâmicos foi realizada e, em várias situações, a proposta deste trabalho superou o desempenho de métodos considerados estado-da-arte / Abstract: The dynamism of the modern world gives rise to huge scientific and technological challenges. Problems until recently being treated as static are now being reformulated to incorporate that dynamism, thus requiring novel solution strategies. Population-based metaheuristics devoted to optimization emerge as promising approaches, given that they promote an effective exploration of the search space and contribute to the adaptation to the dynamism of the environment. Estimation of distribution algorithms (EDAs) were considered here, which make use of probabilistic models to identify promising regions of the search space. Due to the fact that the proposals of EDAs for dynamic problems are rare and limited, mainly in real-parameter search spaces, EDAs were conceived based on flexible Gaussian mixture models, self-controlable and computationally inexpensive steps, including diversity maintenance and convergence control mechanisms. An extensive comparison with alternative optimization methods for dynamic environments was accomplished and, in many situations, the proposed technique overcame the performance produced by state-of-the-art methods / Mestrado / Engenharia de Computação / Mestre em Engenharia Elétrica

[EN] This thesis considers new applications of non-Gaussian mixtures in the framework of statistical signal processing and pattern recognition. The non-Gaussian mixtures were implemented by mixtures of independent component analyzers (ICA). The fundamental hypothesis of ICA is that the observed signals can be expressed as a linear transformation of a set of hidden variables, usually referred to as sources, which are statistically independent. This independence allows factoring the original M-dimensional probability density function (PDF) of the data as a product of one-dimensional probability densities, greatly simplifying the modeling of the data. ICA mixture models (ICAMM) provide further flexibility by alleviating the independency requirement of ICA, thus allowing the model to obtain local projections of the data without compromising its generalization capabilities. Here are explored new possibilities of ICAMM for the purposes of estimation and classification of signals. The thesis makes several contributions to the research in non-Gaussian mixtures: (i) a method for maximum-likelihood estimation of missing data, based on the maximization of the PDF of the data given the ICAMM; (ii) a method for Bayesian estimation of missing data that minimizes the mean squared error and can obtain the confidence interval of the prediction; (iii) a generalization of the sequential dependence model for ICAMM to semi-supervised or supervised learning and multiple chains of dependence, thus allowing the use of multimodal data; and (iv) introduction of ICAMM in diverse novel applications, both for estimation and for classification. The developed methods were validated via an extensive number of simulations that covered multiple scenarios. These tested the sensitivity of the proposed methods with respect to the following parameters: number of values to estimate; kinds of source distributions; correspondence of the data with respect to the assumptions of the model; number of classes in the mixture model; and unsupervised, semi-supervised, and supervised learning. The performance of the proposed methods was evaluated using several figures of merit, and compared with the performance of multiple classical and state-of-the-art techniques for estimation and classification. Aside from the simulations, the methods were also tested on several sets of real data from different types: data from seismic exploration studies; ground penetrating radar surveys; and biomedical data. These data correspond to the following applications: reconstruction of damaged or missing data from ground-penetrating radar surveys of historical walls; reconstruction of damaged or missing data from a seismic exploration survey; reconstruction of artifacted or missing electroencephalographic (EEG) data; diagnosis of sleep disorders; modeling of the brain response during memory tasks; and exploration of EEG data from subjects performing a battery of neuropsychological tests. The obtained results demonstrate the capability of the proposed methods to work on problems with real data. Furthermore, the proposed methods are general-purpose and can be used in many signal processing fields. / [ES] Esta tesis considera nuevas aplicaciones de las mezclas no Gaussianas dentro del marco de trabajo del procesado estadístico de señal y del reconocimiento de patrones. Las mezclas no Gaussianas fueron implementadas mediante mezclas de analizadores de componentes independientes (ICA). La hipótesis fundamental de ICA es que las señales observadas pueden expresarse como una transformación lineal de un grupo de variables ocultas, normalmente llamadas fuentes, que son estadísticamente independientes. Esta independencia permite factorizar la función de densidad de probabilidad (PDF) original M-dimensional de los datos como un producto de densidades unidimensionales, simplificando ampliamente el modelado de los datos. Los modelos de mezclas ICA (ICAMM) aportan una mayor flexibilidad al relajar el requisito de independencia de ICA, permitiendo que el modelo obtenga proyecciones locales de los datos sin comprometer su capacidad de generalización. Aquí se exploran nuevas posibilidades de ICAMM para los propósitos de estimación y clasificación de señales. La tesis realiza varias contribuciones a la investigación en mezclas no Gaussianas: (i) un método de estimación de datos faltantes por máxima verosimilitud, basado en la maximización de la PDF de los datos dado el ICAMM; (ii) un método de estimación Bayesiana de datos faltantes que minimiza el error cuadrático medio y puede obtener el intervalo de confianza de la predicción; (iii) una generalización del modelo de dependencia secuencial de ICAMM para aprendizaje supervisado o semi-supervisado y múltiples cadenas de dependencia, permitiendo así el uso de datos multimodales; y (iv) introducción de ICAMM en varias aplicaciones novedosas, tanto para estimación como para clasificación. Los métodos desarrollados fueron validados mediante un número extenso de simulaciones que cubrieron múltiples escenarios. Éstos comprobaron la sensibilidad de los métodos propuestos con respecto a los siguientes parámetros: número de valores a estimar; tipo de distribuciones de las fuentes; correspondencia de los datos con respecto a las suposiciones del modelo; número de clases en el modelo de mezclas; y aprendizaje supervisado, semi-supervisado y no supervisado. El rendimiento de los métodos propuestos fue evaluado usando varias figuras de mérito, y comparado con el rendimiento de múltiples técnicas clásicas y del estado del arte para estimación y clasificación. Además de las simulaciones, los métodos también fueron probados sobre varios grupos de datos de diferente tipo: datos de estudios de exploración sísmica; exploraciones por radar de penetración terrestre; y datos biomédicos. Estos datos corresponden a las siguientes aplicaciones: reconstrucción de datos dañados o faltantes de exploraciones de radar de penetración terrestre de muros históricos; reconstrucción de datos dañados o faltantes de un estudio de exploración sísmica; reconstrucción de datos electroencefalográficos (EEG) dañados o artefactados; diagnóstico de desórdenes del sueño; modelado de la respuesta del cerebro durante tareas de memoria; y exploración de datos EEG de sujetos durante la realización de una batería de pruebas neuropsicológicas. Los resultados obtenidos demuestran la capacidad de los métodos propuestos para trabajar en problemas con datos reales. Además, los métodos propuestos son de propósito general y pueden utilizarse en muchos campos del procesado de señal. / [CAT] Aquesta tesi considera noves aplicacions de barreges no Gaussianes dins del marc de treball del processament estadístic de senyal i del reconeixement de patrons. Les barreges no Gaussianes van ser implementades mitjançant barreges d'analitzadors de components independents (ICA). La hipòtesi fonamental d'ICA és que els senyals observats poden ser expressats com una transformació lineal d'un grup de variables ocultes, comunament anomenades fonts, que són estadísticament independents. Aquesta independència permet factoritzar la funció de densitat de probabilitat (PDF) original M-dimensional de les dades com un producte de densitats de probabilitat unidimensionals, simplificant àmpliament la modelització de les dades. Els models de barreges ICA (ICAMM) aporten una major flexibilitat en alleugerar el requeriment d'independència d'ICA, permetent així que el model obtinga projeccions locals de les dades sense comprometre la seva capacitat de generalització. Ací s'exploren noves possibilitats d'ICAMM pels propòsits d'estimació i classificació de senyals. Aquesta tesi aporta diverses contribucions a la recerca en barreges no Gaussianes: (i) un mètode d'estimació de dades faltants per màxima versemblança, basat en la maximització de la PDF de les dades donat l'ICAMM; (ii) un mètode d'estimació Bayesiana de dades faltants que minimitza l'error quadràtic mitjà i pot obtenir l'interval de confiança de la predicció; (iii) una generalització del model de dependència seqüencial d'ICAMM per entrenament supervisat o semi-supervisat i múltiples cadenes de dependència, permetent així l'ús de dades multimodals; i (iv) introducció d'ICAMM en diverses noves aplicacions, tant per a estimació com per a classificació. Els mètodes desenvolupats van ser validats mitjançant una extensa quantitat de simulacions que cobriren múltiples situacions. Aquestes van verificar la sensibilitat dels mètodes proposats amb respecte als següents paràmetres: nombre de valors per estimar; mena de distribucions de les fonts; correspondència de les dades amb respecte a les suposicions del model; nombre de classes del model de barreges; i aprenentatge supervisat, semi-supervisat i no-supervisat. El rendiment dels mètodes proposats va ser avaluat mitjançant diverses figures de mèrit, i comparat amb el rendiments de múltiples tècniques clàssiques i de l'estat de l'art per a estimació i classificació. A banda de les simulacions, els mètodes van ser verificats també sobre diversos grups de dades reals de diferents tipus: dades d'estudis d'exploració sísmica; exploracions de radars de penetració de terra; i dades biomèdiques. Aquestes dades corresponen a les següents aplicacions: reconstrucció de dades danyades o faltants d'estudis d'exploracions de radar de penetració de terra sobre murs històrics; reconstrucció de dades danyades o faltants en un estudi d'exploració sísmica; reconstrucció de dades electroencefalogràfiques (EEG) artefactuades o faltants; diagnosi de desordres de la son; modelització de la resposta del cervell durant tasques de memòria; i exploració de dades EEG de subjectes realitzant una bateria de tests neuropsicològics. Els resultats obtinguts han demostrat la capacitat dels mètodes proposats per treballar en problemes amb dades reals. A més, els mètodes proposats són de propòsit general i poden fer-se servir en molts camps del processament de senyal. / Safont Armero, G. (2015). New Insights in Prediction and Dynamic Modeling from Non-Gaussian Mixture Processing Methods [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/53913 / TESIS

Récemment, la grande complexité des applications modernes, par exemple dans la génétique, l’informatique, la finance, les sciences du climat, etc. a conduit à la proposition des nouveaux modèles qui peuvent décrire la réalité. Dans ces cas,méthodes MCMC classiques ne parviennent pas à rapprocher la distribution a posteriori, parce qu’ils sont trop lents pour étudier le space complet du paramètre. Nouveaux algorithmes ont été proposés pour gérer ces situations, où la fonction de vraisemblance est indisponible. Nous allons étudier nombreuses caractéristiques des modèles complexes: comment éliminer les paramètres de nuisance de l’analyse et faire inférence sur les quantités d’intérêt,dans un cadre bayésienne et non bayésienne et comment construire une distribution a priori de référence. / Recently, the great complexity of modern applications, for instance in genetics,computer science, finance, climatic science etc., has led to the proposal of newmodels which may realistically describe the reality. In these cases, classical MCMCmethods fail to approximate the posterior distribution, because they are too slow toinvestigate the full parameter space. New algorithms have been proposed to handlethese situations, where the likelihood function is unavailable. We will investigatemany features of complex models: how to eliminate the nuisance parameters fromthe analysis and make inference on key quantities of interest, both in a Bayesianand not Bayesian setting, and how to build a reference prior.

In der Arbeit wird ein neuartiges Lernverfahren für Gauß-Mischverteilungen entwickelt. Es basiert auf der Technik der Markov-Chain Monte-Carlo Verfahren und ist in der Lage, in einem Zuge die Größe der Mischverteilung sowie deren Parameter zu bestimmen. Das Verfahren zeichnet sich sowohl durch eine gute Anpassung an die Trainingsdaten als auch durch eine gute Generalisierungsleistung aus. Ausgehend von einer Beschreibung der stochastischen Grundlagen und einer Analyse der Probleme, die beim Lernen von Gauß-Mischverteilungen auftreten, wird in der Abeit das neue Lernverfahren schrittweise entwickelt und seine Eigenschaften untersucht. Ein experimenteller Vergleich mit bekannten Lernverfahren für Gauß-Mischverteilungen weist die Eignung des neuen Verfahrens auch empirisch nach.