The process of modeling earthquake hazard risk and vulnerability is a prime component of mitigation planning, but is rife with epistemic, aleatory and factual uncertainty. Reducing uncertainty in such models yields significant benefits, both in terms of extending knowledge and increasing the efficiency and effectiveness of mitigation planning. An accurate description of the built environment as an input into loss estimation would reduce factual uncertainty in the modeling process. Building attributes for earthquake loss estimation and risk assessment modeling were identified. Three modules for developing the building attributes were proposed, including structure classification, building footprint recognition and building valuation. Data from primary sources and field surveys were collected from Shelby County, Tennessee, for calibration and validation of the structure type models and for estimation of various components of building value. Building footprint libraries were generated for implementation of algorithms to programmatically recognize two-dimensional building configurations. The modules were implemented to produce a building inventory for Shelby County, Tennessee that may be used effectively in loss estimation modeling. Validation of the building inventory demonstrates effectively that advanced technologies and methods may be effectively and innovatively applied on combinations of primary and derived data and replicated in order to produce a bottom-up, reliable, accurate and cost-effective building inventory.

Είναι ευρύτατα γνωστό ότι ο θόρυβος δημιουργείται από δονούμενες επιφάνειες. Για την αντιμετώπιση του θορύβου στην πηγή του, δηλαδή τη δονούμενη επιφάνεια, δύο κυρίως τρόποι έχουν αναπτυχθεί. Ο πρώτος τρόπος αφορά τη χρησιμοποίηση παθητικών μέσων, δηλαδή ηχομονωτικών υλικών που αποσβένουν συγκεκριμένες συχνότητες. Ο δεύτερος τρόπος αφορά τη χρήση ενεργητικών μέσων. Τα ενεργητικά μέσα είναι διατάξεις που αποτελούνται από ένα σύστημα ελέγχου και ένα σύνολο αισθητήρων και ενεργοποιητών. Η λειτουργία ενός τέτοιου Συστήματος Ενεργού Ελέγχου Δονήσεων (ΣΕΕΔ) βασίζεται στην καταγραφή μέσω των αισθητήρων του τρόπου δόνησης της επιφάνειας (πρωτεύον πεδίο δόνησης), την δημιουργία σημάτων ελέγχου από τον ελεγκτή (ίδιου πλάτους αλλά με διαφορά φάσης 180o) και την αποστολή τους στους ενεργοποιητές που θα δημιουργήσουν ένα δευτερεύον πεδίο δόνησης. Η υπέρθεση των δύο πεδίων έχει σαν αποτέλεσμα την δημιουργία ενός εναπομείναντος πεδίου με πλάτη δόνησης αισθητά χαμηλότερα από αυτά του πρωτεύοντος. Το αντικείμενο της παρούσας διατριβής είναι η ανάπτυξη ενός γενικευμένου ΣΕΕΔ, ο έλεγχος του οποίου βασίζεται σε εργαλεία Τεχνητής Νοημοσύνης όπως τα Τεχνητά Νευρωνικά Δίκτυα και οι Γενετικοί Αλγόριθμοι για την αναγνώριση του τρόπου δόνησης οποιασδήποτε επιφάνειας και το βέλτιστο έλεγχο της δόνησής της, χωρίς να απαιτείται καμία πρότερη γνώση της δυναμικής συμπεριφοράς της επιφάνειας. Επιπλέον, το υπό μελέτη ΣΕΕΔ είναι ικανό να ελέγχει τέσσερις συχνότητες αντί μιας που απαντάται συνήθως στην πλειονότητα των εφαρμογών. Ο σκοπός της διατριβής αυτής είναι η απόδειξη της αρχής λειτουργίας ενός τέτοιου συστήματος. Η προσέγγιση για την επίτευξη αυτού του στόχου περιλαμβάνει πειραματικές μετρήσεις ενός πρωτότυπου ΣΕΕΔ σε μία απλοποιημένη πειραματική διάταξη. Τα αποτελέσματα από την εφαρμογή του εν λόγω ΣΕΕΔ δείχνουν ότι παρά τους περιορισμούς που υπεισέρχονται λόγω των δυνατοτήτων του υλικού (hardware) του χρησιμοποιούμενου εξοπλισμού, το υπό μελέτη ΣΕΕΔ λειτουργεί επιτυχώς στη βασική αρχή του, ενώ έχει τις προϋποθέσεις και τη δυναμική για περαιτέρω βελτιστοποίηση και εξέλιξη σε ένα ευρύ φάσμα εφαρμογών. / It is generally approved that noise is created by vibrating surfaces. In order to tackle this phenomenon at its source, mainly two approaches have been followed. The first approach involves passive means, that is sound insulation materials that dampen certain frequencies. The second approach involves the use of active means. The active means are arrangements that consist of a control system and a set of sensors and actuators. The application of such an arrangement for vibration control is called Active Vibration Control (AVC) and is based on the sampling (by means of sensors) of the primary field of vibration of the surface, the creation of control signals by the controller (secondary field - of the same amplitude but with phase difference of 180o) and finally applying these control signals on the vibrating surface, by means of the actuators. The superimposing of the two vibration signals (primary and secondary) results to a residual field where the amplitudes of vibration are significantly lower than in the primary. The objective of the thesis at hand is to develop a Generic AVC with the controller developed using Artificial Intelligence tools such as the Artificial Neural Networks (ANNs) and Genetic Algorithms (GAs), in order to identify the vibration patterns of any surface and the optimal control of its vibration, without any prior knowledge of the dynamic behavior of the surface. Moreover, the developed AVC system will be able to identify and control four dominating frequencies instead of one that is usually the choice in the majority of similar applications. The scope of this work is the ‘Proof of Concept’ of the successful operation of such a generic AVC system. The approach to this end includes experimental testing of a prototype AVC system on a simplified experimental set-up. The results of the application of the developed AVC system, performed also by independent parties in the framework of a EC-funded Basic Research project, prove the successful operation of the developed AVCS, even within the limitation of the contemporary data acquisition platform (hardware and software) used, imposes limitations in the efficiency of the AVCS, and provide the basis for its further development and application in a multitude of problems.

Η παρούσα διδακτορική διατριβή πραγματεύεται το θέμα της εκπαίδευσης εμπρόσθιων τροφοδοτούμενων τεχνητών νευρωνικών δικτύων και τις εφαρμογές τους. Η παρουσίαση των θεμάτων και των αποτελεσμάτων της διατριβής οργανώνεται ως εξής: Στο Κεφάλαιο 1 παρουσιάζονται τα τεχνητά νευρωνικά δίκτυα , τα οφέλη της χρήσης τους, η δομή και η λειτουργία τους. Πιο συγκεκριμένα, παρουσιάζεται πως από τους βιολογικούς νευρώνες μοντελοποιούνται οι τεχνητοί νευρώνες, που αποτελούν το θεμελιώδες στοιχείο των τεχνητών νευρωνικών δικτύων. Στη συνέχεια αναφέρονται οι βασικές αρχιτεκτονικές των εμπρόσθιων τροφοδοτούμενων τεχνητών νευρωνικών δικτύων. Το κεφάλαιο ολοκληρώνεται με μια ιστορική αναδρομή για τα τεχνητά νευρωνικά δίκτυα και με την παρουσίαση κάποιων εφαρμογών τους. Στο Κεφάλαιο 2 παρουσιάζονται μερικοί από τους υπάρχοντες αλγορίθμους εκπαίδευσης τεχνητών νευρωνικών δικτύων. Γίνεται μια περιληπτική αναφορά του προβλήματος της εκπαίδευσης των τεχνητών νευρωνικών δικτύων με επίβλεψη και δίνεται η μαθηματική μοντελοποίηση που αντιστοιχεί στην ελαχιστοποίηση του κόστους. Στην συνέχεια γίνεται μια περιληπτική αναφορά στις μεθόδους που βασίζονται στην κατεύθυνση της πιο απότομης καθόδου, στις μεθόδους δευτέρας τάξεως όπου απαιτείται ο υπολογισμός του Εσσιανού πίνακα της συνάρτησης κόστους, στις μεθόδους μεταβλητής μετρικής, και στις μεθόδους συζυγών κλίσεων. Κατόπιν, παρουσιάζεται ο χώρος των βαρών, η επιφάνεια σφάλματος και οι διάφορες τεχνικές αρχικοποίησης των βαρών των τεχνητών νευρωνικών δικτύων και περιγράφονται οι επιπτώσεις που έχουν στην εκπαίδευση τους. Στο Κεφάλαιο 3 παρουσιάζεται ένας νέος αλγόριθμος εκπαίδευσης τεχνητών νευρωνικών δικτύων βασισμένος στον αλγόριθμο της οπισθοδιάδοσης του σφάλματος και στην αυτόματη προσαρμογή του ρυθμού εκπαίδευσης χρησιμοποιώντας πληροφορία δυο σημείων. Η κατεύθυνση αναζήτησης του νέου αλγορίθμου είναι η κατεύθυνση της πιο απότομης καθόδου, αλλά για τον προσδιορισμό του ρυθμού εκπαίδευσης χρησιμοποιούνται προσεγγίσεις δυο σημείων της εξίσωσης χορδής των μεθόδων ψεύδο-Newton. Επιπλέον, παράγεται ένας νέος ρυθμός εκπαίδευσης προσεγγίζοντας την νέα εξίσωση χορδής, που προτάθηκε από τον Zhang, η οποία χρησιμοποιεί πληροφορία παραγώγων και συναρτησιακών τιμών. Στη συνέχεια, ένας κατάλληλος μηχανισμός επιλογής του ρυθμού εκπαίδευσης ενσωματώνεται στον αλγόριθμο εκπαίδευσης ώστε να επιλέγεται κάθε φορά ο κατάλληλος ρυθμός εκπαίδευσης. Τέλος, γίνεται μελέτη της σύγκλισης του αλγορίθμου εκπαίδευσης και παρουσιάζονται τα πειραματικά αποτελέσματα για διάφορα προβλήματα εκπαίδευσης. Στο Κεφάλαιο 4 παρουσιάζονται μερικοί αποτελεσματικοί αλγόριθμοι εκπαίδευσης οι οποίοι βασίζονται στις μεθόδους βελτιστοποίησης συζυγών κλίσεων. Στους υπάρχοντες αλγόριθμους εκπαίδευσης συζυγών κλίσεων προστίθεται ένας αλγόριθμος εκπαίδευσης που βασίζεται στη μέθοδο συζυγών κλίσεων του Perry. Επιπρόσθετα, προτείνονται νέοι αλγόριθμοι συζυγών κλίσεων που προκύπτουν από τις ίδιες αρχές που προέρχονται οι γνωστοί αλγόριθμοι συζυγών κλίσεων των Hestenes-Stiefel, Fletcher-Reeves, Polak-Ribiere και Perry, και ονομάζονται κλιμακωτοί αλγόριθμοι συζυγών κλίσεων. Αυτή η κατηγορία αλγορίθμων βασίζεται στην φασματική παράμετρο κλιμάκωσης του προτάθηκε από τους Barzilai και Borwein. Επιπλέον, ενσωματώνεται στους αλγόριθμους εκπαίδευσης συζυγών κλίσεων μια αποδοτική τεχνική γραμμικής αναζήτησης, που βασίζεται στις συνθήκες του Wolfe και στην διασφαλισμένη κυβική παρεμβολή. Ακόμη, η παράμετρος του αρχικού ρυθμού εκπαίδευσης προσαρμόζεται αυτόματα σε κάθε επανάληψη σύμφωνα με ένα κλειστό τύπο. Στη συνέχεια, εφαρμόζεται μια αποτελεσματική διαδικασία επανεκκίνησης, έτσι ώστε να βελτιωθούν περαιτέρω οι αλγόριθμοι εκπαίδευσης συζυγών κλίσεων και να αποδειχθεί η ολική τους σύγκλιση. Τέλος, παρουσιάζονται τα πειραματικά αποτελέσματα για διάφορα προβλήματα εκπαίδευσης. Στο τελευταίο Κεφάλαιο της παρούσας διδακτορικής διατριβής, απομονώνεται και τροποποιείται ο κλιμακωτός αλγόριθμος του Perry, που παρουσιάστηκε στο προηγούμενο κεφάλαιο. Πιο συγκεκριμένα, ενώ διατηρούνται τα κύρια χαρακτηριστικά του αλγορίθμου εκπαίδευσης, εφαρμόζεται μια διαφορετική τεχνική γραμμικής αναζήτησης η οποία βασίζεται στις μη μονότονες συνθήκες του Wolfe. Επίσης προτείνεται ένας νέος αρχικός ρυθμός εκπαίδευσης για χρήση με τον κλιμακωτό αλγόριθμο εκπαίδευσης συζυγών κλίσεων, ο οποίος φαίνεται να είναι αποδοτικότερος από τον αρχικό ρυθμό εκπαίδευσης που προτάθηκε από τον Shanno όταν χρησιμοποιείται σε συνδυασμό με την μη μονότονη τεχνική γραμμικής αναζήτησης. Στη συνέχεια παρουσιάζονται τα πειραματικά αποτελέσματα για διάφορα προβλήματα εκπαίδευσης. Τέλος, ως εφαρμογή εκπαιδεύεται ένα πολυεπίπεδο εμπρόσθια τροφοδοτούμενο τεχνητό νευρωνικό δίκτυο με τον προτεινόμενο αλγόριθμο για το πρόβλημα της ταξινόμησης καρκινικών κυττάρων του εγκεφάλου και συγκρίνεται η απόδοση του με την απόδοση ενός πιθανοτικού τεχνητού νευρωνικού δικτύου. Η διατριβή ολοκληρώνεται με το Παράρτημα Α’, όπου παρουσιάζονται τα προβλήματα εκπαίδευσης τεχνητών νευρωνικών δικτύων που χρησιμοποιήθηκαν για την αξιολόγηση των προτεινόμενων αλγορίθμων εκπαίδευσης. / In this dissertation the problem of the training of feedforward artificial neural networks and its applications are considered. The presentation of the topics and the results are organized as follows: In the first chapter, the artificial neural networks are introduced. Initially, the benefits of the use of artificial neural networks are presented. In the sequence, the structure and their functionality are presented. More specifically, the derivation of the artificial neurons from the biological ones is presented followed by the presentation of the architecture of the feedforward neural networks. The historical notes and the use of neural networks in real world problems are concluding the first chapter. In Chapter 2, the existing training algorithms for the feedforward neural networks are considered. First, a summary of the training problem and its mathematical formulation, that corresponds to the uncostrained minimization of a cost function, are given. In the sequence, training algorithms based on the steepest descent, Newton, variable metric and conjugate gradient methods are presented. Furthermore, the weight space, the error surface and the techniques of the initialization of the weights are described. Their influence in the training procedure is discussed. In Chapter 3, a new training algorithm for feedforward neural networks based on the backpropagation algorithm and the automatic two-point step size (learning rate) is presented. The algorithm uses the steepest descent search direction while the learning rate parameter is calculated by minimizing the standard secant equation. Furthermore, a new learning rate parameter is derived by minimizing the modified secant equation introduced by Zhang, that uses both gradient and function value information. In the sequece a switching mechanism is incorporated into the algorithm so that the appropriate stepsize to be chosen according to the status of the current iterative point. Finaly, the global convergence of the proposed algorithm is studied and the results of some numerical experiments are presented. In Chapter 4, some efficient training algorithms, based on conjugate gradient optimization methods, are presented. In addition to the existing conjugate gradient training algorithms, we introduce Perry's conjugate gradient method as a training algorithm. Furthermore, a new class of conjugate gradient methods is proposed, called self-scaled conjugate gradient methods, which are derived from the principles of Hestenes-Stiefel, Fletcher-Reeves, Polak-Ribiere and Perry's method. This class is based on the spectral scaling parameter. Furthermore, we incorporate to the conjugate gradient training algorithms an efficient line search technique based on the Wolfe conditions and on safeguarded cubic interpolation. In addition, the initial learning rate parameter, fed to the line search technique, was automatically adapted at each iteration by a closed formula. Finally, an efficient restarting procedure was employed in order to further improve the effectiveness of the conjugate gradient training algorithms and prove their global convergence. Experimental results show that, in general, the new class of methods can perform better with a much lower computational cost and better success performance. In the last chapter of this dissertation, the Perry's self-scaled conjugate gradient training algorithm that was presented in the previous chapter was isolated and modified. More specifically, the main characteristics of the training algorithm were maintained but in this case a different line search strategy based on the nonmonotone Wolfe conditions was utilized. Furthermore, a new initial learning rate parameter was introduced for use in conjunction with the self-scaled conjugate gradient training algorithm that seems to be more effective from the initial learning rate parameter, proposed by Shanno, when used with the nonmonotone line search technique. In the sequence the experimental results for differrent training problems are presented. Finally, a feedforward neural network with the proposed algorithm for the problem of brain astrocytomas grading was trained and compared the results with those achieved by a probabilistic neural network. The dissertation is concluded with the Appendix A', where the training problems used for the evaluation of the proposed training algorithms are presented.

Nos últimos anos, as redes de comunicação móveis se tornaram de fundamental importância para a infraestrutura dos sistemas de comunicação. Uma das áreas de maior crescimento é a computação móvel. Realizada através de sinais de rádio, a quantidade de canais disponíveis raramente é suficiente para atender a crescente demanda. Este trabalho apresenta uma solução para a questão da alocação de canais, um tópico desafiador dentro da área de redes móveis. A implementação de alocação dinâmica com uso de técnicas computacionais clássicas melhora a utilização dos recursos disponíveis,mas necessita de ajustes periódicos para se adequar a novos cenários. Para a construção de um sistema mais flexível e adaptável, a abordagem escolhida utiliza técnicas de Inteligência Artificial. O modelo proposto combina Teoria Nebulosa, Redes Neurais Artificiais e Sistemas Multi-Agentes. As características de cada técnica foram analisadas e identificamos as partes do sistema que poderiam ser beneficiadas por cada uma. O sistema é resultado da combinação coordenada das três técnicas, e constitui um método eficiente e flexível para gerenciamento de recursos de rádio. Após o detalhamento do modelo, realizamos uma simulação de uma rede celular com o sistema proposto e seu comportamento é comparado com uma rede de referência, para verificação das diferenças e melhorias alcançadas. Por fim, apresentamos a situação atual da pesquisa e os possíveis caminhos para aprimoramento do sistema. / In the last years, mobile networks became more important for communication systems’ infrastructure. One area of great growth is mobile computation, which is performed through radio signals. The amount of available channels rarely is enough to attend the increasing demand. This work presents a solution for the channel allocation topic, a challenging topic inside mobile networks area. The implementation of dynamic allocation using classic computational techniques improves the use of available resources, but it needs periodic and frequent adjustments for new scenarios. The construction of a more flexible and adaptable system was achieved using Artificial Intelligence techniques. Proposed model combines Fuzzy Logic, Artificial Neural Networks and Multi-Agents Systems. Features of each technique had been analyzed and we identified the system modules which could be benefited by them. The system is the result of coordinated combination of these three techniques, and constitutes an efficient and flexible method for radio resources management. After model detailing, we executed a cellular network simulation using proposed system, and its behavior is compared with a reference network, presenting reached differences and improvements. Finally, we present current situation of this research and possible ways for system improvement.

Devido a restrições de aplicabilidade presentes nos algoritmos para a solução de problemas de otimização combinatória, os sistemas baseados em redes neurais artificiais e algoritmos genéticos oferecem um método alternativo para solucionar tais problemas eficientemente. Os algoritmos genéticos devem a sua popularidade à possibilidade de percorrer espaços de busca não-lineares e extensos. Já as redes neurais artificiais possuem altas taxas de processamento por utilizarem um número elevado de elementos processadores simples com alta conectividade entre si. Complementarmente, redes neurais com conexões realimentadas fornecem um modelo computacional capaz de resolver vários tipos de problemas de otimização, os quais consistem, geralmente, da otimização de uma função objetivo que pode estar sujeita ou não a um conjunto de restrições. Esta tese apresenta uma abordagem inovadora para resolver problemas de conexão em otimização combinatória utilizando uma arquitetura neuro-genética. Mais especificamente, uma rede neural de Hopfield modificada é associada a um algoritmo genético visando garantir a convergência da rede em direção aos pontos de equilíbrio factíveis que representam as soluções para os problemas de otimização combinatória. / Due to applicability constraints involved with the algorithms for solving combinatorial optimization problems, systems based on artificial neural networks and genetic algorithms are alternative methods for solving these problems in an efficient way. The genetic algorithms must its popularity to make possible cover nonlinear and extensive search spaces. On the other hand, artificial neural networks have high processing rates due to the use of a massive number of simple processing elements and the high degree of connectivity between these elements. Additionally, neural networks with feedback connections provide a computing model capable of solving a large class of optimization problems, which refer to optimization of an objective function that can be subject to constraints. This thesis presents a novel approach for solving connection problems in combinatorial optimization using a neurogenetic approach. More specifically, a modified Hopfield neural network is associated with a genetic algorithm in order to guarantee the convergence of the network to the equilibrium points, which represent feasible solutions for the combinatorial optimization problems.

O presente trabalho é resultado da colaboração de pesquisadores do Centro de Engenharia Nuclear (CEN) e de pesquisadores do Centro de Biotecnologia (CB), ambos pertencentes ao IPEN, para o desenvolvimento de uma metodologia que visa auxiliar os profissionais citogeneticistas fornecendo uma ferramenta que automatize parte da rotina necessária para a avaliação qualitativa e quantitativa de danos biológicos em termos de aberração cromossômica. A técnica citogenética, sobre a qual esta ferramenta é desenvolvida, é a técnica de aberrações cromossômicas. Nela, são realizadas preparações citológicas de linfócitos de sangue periférico para que metáfases sejam analisadas e fotografadas ao microscópio e, com base na morfologia dos cromossomos, anomalias sejam investigadas. Quando esta tarefa é realizada manualmente, os cromossomos são analisados visualmente um a um pelo profissional citogeneticista, logo, trata-se de um processo minucioso em virtude da variação geral na aparência do cromossomo, do seu tamanho pequeno e do grande número de cromossomos por célula. Para um diagnóstico confiável, é necessário que várias células sejam analisadas, tornando-se uma tarefa repetitiva e demorada. Neste contexto, foi proposto o uso dos mapas auto-organizáveis para o reconhecimento automático de padrões morfológicos referentes às imagens de cromossomos humanos. Para isso, foi desenvolvido um método de extração de características por meio do qual é possível classificar os cromossomos em: dicêntricos, anéis, acrocêntricos, submetacêntricos e metacêntricos, com acerto de 93,4 % em relação ao diagnóstico dado por um profissional citogeneticista. / This work is a joint collaboration between Nuclear Energy Research Institute (IPEN), Nuclear Engineering Center and Biotechnology Center to develop a methodology aiming to assist cytogenetic professionals by providing a tool to automate part of the required routine to perform qualitative and quantitative evaluation of biological damage in terms of chromosomal aberration. The cytogenetic technique upon which this tool was developed, is the chromosome aberrations technique, in which cytological preparations of peripheral blood lymphocyte metaphases are performed to be analyzed and photographed under a microscope in order to investigating chromosomal aberration. Performed manually, the chromosomes are analyzed visually one by one by a cytogenetic professional, so it is a painstaking process due to the great deal of variation in the appearance of each chromosome, their small sizes and not to mention the high density of chromosomes per cell. In order to obtain a reliable diagnosis it is necessary that many cells be analyzed, which makes this a repetitive and time consuming process. In this context, the use of self-organizing maps for the automatic recognition of patterns relating to morphological pictures of human chromosomes has been proposed. For this, we developed a feature extraction method by which is possible to classify chromosomes in: dicentrics, ring-shaped, acrocentric, submetacentric and metacentric with 93.4% accuracy compared to diagnostic given by a professional cytogeneticist.

Made available in DSpace on 2014-12-17T15:47:47Z (GMT). No. of bitstreams: 1 ArakenMS_da_capa_ate_pag_66.pdf: 612002 bytes, checksum: 77ee53e5ec8496b7cf1c4503e222c41d (MD5) Previous issue date: 2008-02-01 / RePART (Reward/Punishment ART) is a neural model that constitutes a variation of the Fuzzy Artmap model. This network was proposed in order to minimize the inherent problems in the Artmap-based model, such as the proliferation of categories and misclassification. RePART makes use of additional mechanisms, such as an instance counting parameter, a reward/punishment process and a variable vigilance parameter. The instance counting parameter, for instance, aims to minimize the misclassification problem, which is a consequence of the sensitivity to the noises, frequently presents in Artmap-based models. On the other hand, the use of the variable vigilance parameter tries to smoouth out the category proliferation problem, which is inherent of Artmap-based models, decreasing the complexity of the net. RePART was originally proposed in order to minimize the aforementioned problems and it was shown to have better performance (higer accuracy and lower complexity) than Artmap-based models. This work proposes an investigation of the performance of the RePART model in classifier ensembles. Different sizes, learning strategies and structures will be used in this investigation. As a result of this investigation, it is aimed to define the main advantages and drawbacks of this model, when used as a component in classifier ensembles. This can provide a broader foundation for the use of RePART in other pattern recognition applications / O RePART (Reward/Punishiment ART), modelo neural que se constitui numa varia??o do modelo Fuzzy Artmap, foi proposto objetivando minimizar problemas inerentes aos modelos da classe Artmap, tais como: prolifera??o de categorias e m? classifica??o. Por essa raz?o, o RePART faz uso de mecanismos adicionais, como: um par?metro contador de inst?ncia, um processo de recompensa/puni??o e um par?metro de vigil?ncia vari?vel. O par?metro contador de inst?ncia busca minimizar o problema de m? classifica??o, resultante da sensibilidade ? ru?dos, freq?entemente presente nos modelos da classe Artmap. O uso da vigil?ncia vari?vel tem como objetivo minimizar o problema de prolifera??o de categorias, diminuindo a complexidade da rede, quando utilizado em aplica??es com um grande n?mero de padr?es de treinamento. A proposta do RePART visou a minimiza??o desses problemas e foi mostrado que o RePART obteve desempenho superior que alguns modelos da classe Artmap. Neste trabalho ? proposta a realiza??o de uma investiga??o do desempenho do modelo RePART em comit?s de classificadores. Nesta investiga??o ser? realizada uma an?lise com comit?s utilizando diferentes tamanhos, estrat?gias de aprendizados e estruturas. Os resultados obtidos com esta investiga??o servir?o como meio de descoberta das vantagens e desvantagens de cada um dos modelos abordados em comit?s. Com isso, poder? ser dado um embasamento ainda mais amplo ? utiliza??o do RePART em outras aplica??es de reconhecimento de padr?es

Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro / O objetivo deste trabalho é conhecer e compreender melhor os imprevistos no fornecimento de energia elétrica, quando ocorrem as variações de tensão de curta duração (VTCD). O banco de dados necessário para os diagnósticos das faltas foi obtido através de simulações de um modelo de alimentador radial através do software PSCAD/EMTDC. Este trabalho utiliza um Phase-Locked Loop (PLL) com o intuito de detectar VTCDs e realizar a estimativa automática da frequência, do ângulo de fase e da amplitude das tensões e correntes da rede elétrica. Nesta pesquisa, desenvolveram-se duas redes neurais artificiais: uma para identificar e outra para localizar as VTCDs ocorridas no sistema de distribuição de energia elétrica. A técnica aqui proposta aplica-se a alimentadores trifásicos com cargas desequilibradas, que podem possuir ramais laterais trifásicos, bifásicos e monofásicos. No desenvolvimento da mesma, considera-se que há disponibilidade de medições de tensões e correntes no nó inicial do alimentador e também em alguns pontos esparsos ao longo do alimentador de distribuição. Os desempenhos das arquiteturas das redes neurais foram satisfatórios e demonstram a viabilidade das RNAs na obtenção das generalizações que habilitam o sistema para realizar a classificação de curtos-circuitos. / The objective of this work is to know and understand the unforeseen in the supply of electricity, when there are short duration voltage variations (SDVV). The required databases for the diagnosis of faults were obtained through simulations of a model of radial feeder through software PSCAD/EMTDC. This work uses a Phase-Locked Loop (PLL) in order to detect and perform the estimation SDVV automatic frequency, phase angle and amplitude of the voltage and current from the power grid. This research is developing two artificial neural networks: one to identify and another to locate the SDVV occurred in the distribution system of electricity. The technique proposed here applies to three-phase feeders with unbalanced loads, which can have side extensions triphasic, biphasic and monophasic. In developing the same, it is considered that there is availability of measurements of voltages and currents at the node of the initial feeder and also in some points scattered along the distribution feeder. The performances of the architectures of neural networks were satisfactory and demonstrate the feasibility of ANNs in obtaining the generalizations that enables the system for the classification of short circuits.

Studiens syfte är att öka förståelsen för möjligheterna med modern artificiell intelligens (AI) vid militär användning genom att bidra med information om ny teknik. Moderna civila applikationer av datorseende som skapats genom användande av så kallade artificiella neurala nät visar resultat som närmar sig den mänskliga synens nivå när det gäller att känna igen olika saker i sin omgivning. Denna studie motiveras av dessa observationer inom området AI i förhållande till situationer i luftstrid då pilotens syn används för att känna igen flygplan innan det bekämpas. Exempelvis vid användande av hjälmsikte riktar pilotens ögon målsökaren hos en luftmålsrobot mot det flygplan som robotens målsökare sedan låser på. Utanför visuella avstånd kan pilotens ögon av naturliga skäl inte användas på detta sätt, varför datorseende använt i en luftmålsrobot undersöks. Resultaten från studien stödjer att datorseende genom användande av artificiella neurala nät kan användas i en luftmålsrobot samt att datorseende kan utföra uppgiften att känna igen stridsflygplan. / This study is aimed at increasing the knowledge to those concerned within the Armed Forces by providing information on the possibilities of modern artificial intelligence (AI). Motivation comes from observations of civilian technology on the use of AI in the field of Computer Vision showing performance equaling the level of the human vision when using the technology of Deep Learning of Artificial Neural Nets. In air-combat aircraft the pilot´s vision is used for recognizing the aircraft that is about to be shot down. For example when utilizing helmet mounted displays, the seeker of an air-target-missile is directed upon the aircraft on which the pilot´s eyes are looking. However when air-target-missiles are utilized beyond visual range the pilot´s vision cannot help in directing the seeker on a specific target. Therefore computer vision within an air-target-missile is studied. The results of the study support that the technology of neural networks may be used in an air-target-missile and that computer vision provided by this technology can do the job of recognizing a combat aircraft.

Notwithstanding the recent successes of deep convolutional neural networks for classification tasks, they are sensitive to the selection of their hyperparameters, which impose an exponentially large search space on modern convolutional models. Traditional hyperparameter selection methods include manual, grid, or random search, but these require expert knowledge or are computationally burdensome. Divergently, Bayesian optimization and evolutionary inspired techniques have surfaced as viable alternatives to the hyperparameter problem. Thus, an alternative hybrid approach that combines the advantages of these techniques is proposed. Specifically, the search space is partitioned into discrete-architectural, and continuous and categorical hyperparameter subspaces, which are respectively traversed by a stochastic genetic search, followed by a genetic-Bayesian search. Simulations on a prominent image classification task reveal that the proposed method results in an overall classification accuracy improvement of 0.87% over unoptimized baselines, and a greater than 97% reduction in computational costs compared to a commonly employed brute force approach. / Electrical and Mining Engineering / M. Tech. (Electrical Engineering)