Este trabalho fornece uma avaliação empírica do desempenho de Redes Bayesianas quando se impõe restrições à largura de árvore de sua estrutura. O desempenho da rede é visto especificamente pela sua capacidade de generalização e também pela precisão da inferência em problemas de tomada de decisão. Resultados preliminares sugerem que adicionar essa restrição na largura de árvore diminui a capacidade de generalização do modelo além de tornar a tarefa de aprendizado mais difícil. / This work provides an empirical evaluation of the performance of Bayesian Networks when treewidth is bounded. The performance of the network is viewed as its generalizability and also as the accuracy of inference in decision making problems. Preliminary results suggest that adding constraints to treewidth decreases the model performance on unseen data and makes the corresponding optimization problem more difficult.

Deep learning has in recent years come to dominate the previously separate fields of research in machine learning, computer vision, natural language understanding and speech recognition. Despite breakthroughs in training deep networks, there remains a lack of understanding of both the optimization and structure of deep networks. The approach advocated by many researchers in the field has been to train monolithic networks with excess complexity, and strong regularization --- an approach that leaves much to desire in efficiency. Instead we propose that carefully designing networks in consideration of our prior knowledge of the task and learned representation can improve the memory and compute efficiency of state-of-the art networks, and even improve generalization --- what we propose to denote as structural priors. We present two such novel structural priors for convolutional neural networks, and evaluate them in state-of-the-art image classification CNN architectures. The first of these methods proposes to exploit our knowledge of the low-rank nature of most filters learned for natural images by structuring a deep network to learn a collection of mostly small, low-rank, filters. The second addresses the filter/channel extents of convolutional filters, by learning filters with limited channel extents. The size of these channel-wise basis filters increases with the depth of the model, giving a novel sparse connection structure that resembles a tree root. Both methods are found to improve the generalization of these architectures while also decreasing the size and increasing the efficiency of their training and test-time computation. Finally, we present work towards conditional computation in deep neural networks, moving towards a method of automatically learning structural priors in deep networks. We propose a new discriminative learning model, conditional networks, that jointly exploit the accurate representation learning capabilities of deep neural networks with the efficient conditional computation of decision trees. Conditional networks yield smaller models, and offer test-time flexibility in the trade-off of computation vs. accuracy.

Este trabalho fornece uma avaliação empírica do desempenho de Redes Bayesianas quando se impõe restrições à largura de árvore de sua estrutura. O desempenho da rede é visto especificamente pela sua capacidade de generalização e também pela precisão da inferência em problemas de tomada de decisão. Resultados preliminares sugerem que adicionar essa restrição na largura de árvore diminui a capacidade de generalização do modelo além de tornar a tarefa de aprendizado mais difícil. / This work provides an empirical evaluation of the performance of Bayesian Networks when treewidth is bounded. The performance of the network is viewed as its generalizability and also as the accuracy of inference in decision making problems. Preliminary results suggest that adding constraints to treewidth decreases the model performance on unseen data and makes the corresponding optimization problem more difficult.

Made available in DSpace on 2015-10-19T11:49:44Z (GMT). No. of bitstreams: 1 gustfreire2004.pdf: 813616 bytes, checksum: 86f5b33ac9df89049239d69315302ba8 (MD5) Previous issue date: 2004-03-23 / Contemporary society is based on two pillars: information and knowledge, supported by digital technologies. It brings the necessity of a new attitude and the acquisition of new skills by the users, mainly in the process of communicating the information. This new attitudes and skills result in a search for continuous learning and in the use of intellectual technologies. This process occurs in every single level of contemporary society, involving activities such as training and capability improvement and, mainly, production and management of information. In this sense, learning virtual networks are fundamental to facilitate the communication of information in a society whose structure is becoming more and more distant of hierarchy. This digital network is presented in a new channel of communication of information: the cyberspace. In this process, professors and professionals of information are seen as facilitation agents; on the other side, in learning virtual networks, a professional of information can also be a manager of virtual environment / A sociedade contemporânea tem na informação e no conhecimento os seus pilares, sustentados pelas tecnologias digitais. Para os usuários, isto implica em novas atitudes e na aquisição de novas competências, principalmente no processo de comunicação de informação. Por sua vez, estas novas atitudes e competências resultam em uma necessidade de aprendizado contínuo e no uso de tecnologias intelectuais. Este processo ocorre em todos os níveis da sociedade, envolvendo atividades de treinamento e capacitação e, principalmente, produção e gestão de informação. Neste sentido, as redes virtuais de aprendizagem são fundamentais para facilitar a comunicação de informação em uma sociedade que se estrutura cada vez mais de forma nãohierarquizada. Estas redes digitais se apresentam em um novo canal de comunicação de informação: o ciberespaço. Nesse processo, os professores e profissionais de informação são vistos como facilitadores, sendo que o profissional da informação pode atuar também como um gestor de redes de comunicação da informação em ambiente virtual

<div>U-net is an image segmentation technique developed primarily for medical image analysis that can precisely segment images using a scarce amount of training data. These traits provide U-net with a high utility within the medical imaging community and have resulted in extensive adoption of U-net as the primary tool for segmentation tasks in medical imaging. The success of U-net is evident in its widespread use in nearly all major image modalities from CT scans and MRI to X-rays and microscopy. Furthermore, while U-net is largely a segmentation tool, there have been instances of the use of U-net in other applications. Given that U-net's potential is still increasing, this review examines the numerous developments and breakthroughs in the U-net architecture and provides observations on recent trends. We also discuss the many innovations that have advanced in deep learning and discuss how these tools facilitate U-net. In addition, we review the different image modalities and application areas that have been enhanced by U-net.</div><div>In recent years, deep learning for health care is rapidly infiltrating and transforming medical fields thanks to the advances in computing power, data availability, and algorithm development. In particular, U-Net, a deep learning technique, has achieved remarkable success in medical image segmentation and has become one of the premier tools in this area. While the accomplishments of U-Net and other deep learning algorithms are evident, there still exist many challenges in medical image processing to achieve human-like performance. In this thesis, we propose a U-net architecture that integrates a residual skip connections and recurrent feedback with EfficientNet as a pretrained encoder. Residual connections help feature propagation in deep neural networks and significantly improve performance against networks with a similar number of parameters while recurrent connections ameliorate gradient learning. We also propose a second model that utilizes densely connected layers aiding deeper neural networks. And the proposed third model that incorporates fractal expansions to bypass diminishing gradients. EfficientNet is a family of powerful pretrained encoders that streamline neural network design. The use of EfficientNet as an encoder provides the network with robust feature extraction that can be used by the U-Net decoder to create highly accurate segmentation maps. The proposed networks are evaluated against state-of-the-art deep learning based segmentation techniques to demonstrate their superior performance.</div>

This master thesis is discussing application of reinforcement learning in deep learning tasks. In theoretical part, basics about artificial neural networks and reinforcement learning. The thesis describes theoretical model of reinforcement learning process - Markov processes. Some interesting techniques are shown on conventional reinforcement learning algorithms. Some of widely used deep reinforcement learning algorithms are described here as well. Practical part consist of implementing model of robot and it's environment and of the deep reinforcement learning system itself.

No description available.

Die vorliegende Arbeit widmet sich der Untersuchung und Weiterentwicklung selbständig lernender maschineller Lernverfahren (Reinforcement Lernen) in der Anwendung auf visuelle Wahrnehmungen. Zuletzt wurden mit der Einführung speicherbasierter Methoden in das Reinforcement Lernen große Fortschritte beim Lernen an realen Systemen erzielt, aber der Umgang mit hochkomplexen visuellen Eingabedaten, wie sie z.B. von einer digitalen Kamera aufgezeichnet werden, stellt weiterhin ein ungelöstes Problem dar. Bestehende Methoden sind auf den Umgang mit niedrigdimensionalen Zustandsbeschreibungen beschränkt, was eine Anwendung dieser Verfahren direkt auf den Strom von Bilddaten bisher ausschließt und den vorgeschalteten Einsatz klassischer Methoden des Bildverstehens zur Extraktion und geeigneten Kodierung der relevanten Informationen erfordert. Einen Ausweg bietet der Einsatz von so genannten `tiefen Autoencodern'. Diese mehrschichtigen neuronalen Netze ermöglichen es, selbstorganisiert niedrigdimensionale Merkmalsräume zur Repräsentation hochdimensionaler Eingabedaten zu erlernen und so eine klassische, aufgabenspezifische Bildanalyse zu ersetzen. In typischen Objekterkennungsaufgaben konnten auf Basis dieser erlernten Repräsentationen bereits beeindruckende Ergebnisse erzielt werden. Im Rahmen der vorliegenden Arbeit werden nun die tiefen Autoencodernetze auf ihre grundsätzliche Tauglichkeit zum Einsatz im Reinforcement Lernen untersucht. Mit dem ``Deep Fitted Q''-Algorithmus wird ein neuer Algorithmus entwickelt, der das Training der tiefen Autoencodernetze auf effiziente Weise in den Reinforcement Lernablauf integriert und so den Umgang mit visuellen Wahrnehmungen beim Strategielernen ermöglicht. Besonderes Augenmerk wird neben der Dateneffizienz auf die Stabilität des Verfahrens gelegt. Im Anschluss an eine Diskussion der theoretischen Aspekte des Verfahrens wird eine ausführliche empirische Evaluation der erzeugten Merkmalsräume und der erlernten Strategien an simulierten und realen Systemen durchgeführt. Dabei gelingt es im Rahmen der vorliegenden Arbeit mit Hilfe der entwickelten Methoden erstmalig, Strategien zur Steuerung realer Systeme direkt auf Basis der unvorverarbeiteten Bildinformationen zu erlernen, wobei von außen nur das zu erreichende Ziel vorgegeben werden muss.

Cloud Computing sees increased adoption as companies seek to increase flexibility and reduce cost. Although the large cloud service providers employ a pay-as-you-go pricing model and enable customers to scale up and down quickly, there is still room for improvement. Workload in the form of CPU utilization often fluctuates which leads to unnecessary cost and environmental impact for companies. To help mitigate this issue, the aim of this paper is to predict future CPU utilization using a long short-term memory (LSTM) machine learning model. By predicting utilization up to 30 minutes into the future, companies are able to scale their capacity just in time and avoid unnecessary cost and damage to the environment. The study is divided into two parts. The first part analyses how well the LSTM model performs when predicting one step at a time compared with a state-of-the-art model. The second part analyses the accuracy of the LSTM when making predictions up to 30 minutes into the future. To allow for an objective analysis of results, the LSTM is compared with a standard RNN, which is similar to the LSTM in its inherit algorithmic structure. To conclude, the results suggest that LSTM may be a useful tool for reducing cost and unnecessary environmental impact for business applications hosted on a public cloud. / Användandet av molntjänster ökar bland företag som önskar förbättrad flexibilitet och sänkta kostnader. De stora molntjänstleverantörerna använder en prismodell där kostnaden är direkt kopplad till användningen, och låter kunderna snabbt ställa om sin kapacitet, men det finns ändå förbättringsmöjligheter. CPU-behoven fluktuerar ofta vilket leder till meningslösa kostnader och onödig påverkan på klimatet när kapacitet är outnyttjad. För att lindra detta problem används i denna rapport en LSTM maskininlärningsmodell för att förutspå framtida CPU-utnyttjande. Genom att förutspå utnyttjandet upp till 30 minuter in i framtiden hinner företag ställa om sin kapacitet och undvika onödig kostnad och klimatpåverkan. Arbetet ¨ar uppdelat i två delar. Först en del där LSTM-modellen förutspår ett tidssteg åt gången. Därefter en del som analyserar träffsäkerheten för LSTM flera tidssteg in i framtiden, upp till 30 tidssteg. För att möjliggöra en objektiv utvärdering så jämfördes LSTM-modellen med ett standard recurrent neural network (RNN) vilken liknar LSTM i sin struktur. Resultaten i denna studie visar att LSTM verkar vara ¨överlägsen RNN, både när det gäller att förutspå ett tidssteg in i framtiden och när det gäller flera tidssteg in i framtiden. LSTM-modellen var kapabel att förutspå CPU-utnyttjandet 30 minuter in i framtiden med i hög grad bibehållen träffsäkerhet, vilket också var målet med studien. Sammanfattningsvis tyder resultaten på att denna LSTM-modell, och möjligen liknande LSTM-modeller, har potential att användas i samband med företagsapplikationer då man önskar att reducera onödig kostnad och klimatpåverkan.

La sociedad actual demanda cada vez más aplicaciones computacionalmente exigentes y que se implementen de forma energéticamente eficiente. Esto obliga a la industria del semiconductor a mantener una continua progresión de la tecnología CMOS. No obstante, los expertos vaticinan que el fin de la era de la progresión de la tecnología CMOS se acerca, puesto que se prevé que alrededor del 2020 la tecnología CMOS llegue a su límite. Cuando ésta llegue al punto conocido como “Red Brick Wall”, las limitaciones físicas, tecnológicas y económicas no harán viable el proseguir por esta senda. Todo ello ha motivado que a lo largo de la última década tanto instituciones públicas como privadas apostasen por el desarrollo de soluciones tecnológicas alternativas como es el caso de la nanotecnología (nanotubos, nanohilos, tecnologías basadas en el grafeno, etc.). En esta tesis planteamos una solución alternativa para poder afrontar algunos de los problemas computacionalmente exigentes. Esta solución hace uso de la tecnología CMOS actual sustituyendo la forma de computación clásica desarrollada por Von Neumann por formas de computación no convencionales. Éste es el caso de las computaciones basadas en lógicas pulsantes y en especial la conocida como computación estocástica, la cual proporciona un aumento de la fiabilidad y del paralelismo en los sistemas digitales. En esta tesis se presenta el desarrollo y evaluación de todo un conjunto de bloques computacionales estocásticos implementados mediante elementos digitales clásicos. A partir de estos bloques se proponen diversas metodologías computacionalmente eficientes que mediante su uso permiten afrontar algunos problemas de computación masiva de forma mucho más eficiente. En especial se ha centrado el estudio en los problemas relacionados con el campo del reconocimiento de patrones. / Today's society demands the use of applications with a high computational complexity that must be executed in an energy-efficient way. Therefore the semiconductor industry is forced to maintain the CMOS technology progression. However, experts predict that the end of the age of CMOS technology progression is approaching. It is expected that at 2020 CMOS technology would reach the point known as "Red Brick Wall" at which the physical, technological and economic limitations of CMOS technology will be unavoidable. All of this has caused that over the last decade public and private institutions has bet by the development of alternative technological solutions as is the case of nanotechnology (nanotubes, nanowires, graphene, etc.). In this thesis we propose an alternative solution to address some of the computationally exigent problems by using the current CMOS technology but replacing the classical computing way developed by Von Neumann by other forms of unconventional computing. This is the case of computing based on pulsed logic and especially the stochastic computing that provide a significant increase of the parallelism and the reliability of the systems. This thesis presents the development and evaluation of different stochastic computing methodologies implemented by digital gates. The different methods proposed are able to face some massive computing problems more efficiently than classical digital electronics. This is the case of those fields related to pattern recognition, which is the field we have focused the main part of the research work developed in this thesis.