This thesis deals with a complex analysis and prediction of foreign exchange markets. It uses advanced artificial intelligence methods, namely neural networks and chaos theory. It introduces unconventional approaches and methods of each of these areas, compares them and uses on a real problem. The core of this thesis is a comparison of several prediction models based on completely different principles and underlying theories. The outcome is then a selection of the most appropriate prediction model called NAR + H. The model is evaluated according to several criteria, the pros and cons are discussed and approximate expected profitability and risk are calculated. All analytical, prediction and partial algorithms are implemented in Matlab development environment and form a unified library of all used functions and scripts. It also may be considered as a secondary main outcome of the thesis.

Avec l’avènement de l’ère intellectuelle et de l’Internet of Everything (IoE), les besoins de la communication mondiale et des applications diverses ont explosé. Cette révolution exige que les futurs réseaux de communication soient plus efficaces, intellectuels, agiles et évolutifs. De nombreuses technologies réseau sont apparues pour répondre à la tendance des réseaux de communication de nouvelle génération tels que les réseaux optiques élastiques (EONs) et la virtualisation de réseau. De nombreux défis apparaissent avec les apparences de la nouvelle architecture et de la nouvelle technologie, telles que le routage et l’allocation de ressource spectrale (RSA) dans les EONs et l’intégration de réseaux virtuels (Virtual Network Embedding ou VNE) dans la virtualisation de réseau.Cette thèse traite la conception et l’analyse d’algorithmes d’approximation dans trois problèmes d’optimation du RSA et du VNE : les impacts de la distribution du trafic et de la topologie du réseau sur le routage tout optique, de l’allocation de ressource spectrale, et du VNE dans les topologies des chemins et cycles. Pour le routage tout optique, le premier sous-problème du RSA, il y a toujours un problème en suspens concernant l’impact de la distribution du trafic et de la topologie EON. Comme le routage tout optique joue un rôle essentiel pour la performance globale de la RSA, cette thèse fournit une analyse approfondi théorique sur ces impacts. Pour le deuxième sous-problème du RSA, l’allocation de ressource spectrale, deux chemins optiques quelconques partageant des fibres optiques communes pourraient devoir être isolés dans le domaine spectral avec une bande de garde appropriée pour empêcher la diaphonie et / ou réduire les menaces de sécurité de la couche physique. Cette thèse considère le scénario dans lequel les exigences de bandes de garde réelles optiques sont différentes pour différentes paires de chemins, et étudie comment affecter les ressources spectrales efficacement dans une telle situation. L’hétérogénéité de la topologie des demandes de réseau virtuel (VNR) est un facteur important qui entrave les performances de la VNE. Cependant, dans de nombreuses applications spécialisées, les VNR ont des caractéristiques structurelles communes par exemple, des chemins et des cycles. Pour obtenir de meilleurs résultats, il est donc essentiel de concevoir des algorithmes dédiés pour ces applications en tenant compte des caractéristiques topologiques. Dans cette thèse, nous prouvons que les problèmes VNE dans les topologies de chemin et de cycle sont NP-difficiles. Afin de les résoudre, nous proposons des algorithmes efficaces également analysons leurs ratios d’approximation / With the coming of intellectual era and Internet of Everything (IoE), the needs of worldwide communication and diverse applications have been explosively growing. This information revolution requires the future communication networks to be more efficient, intellectual, agile and scalable. Many technologies have emerged to meet the requirements of next generation communication networks such as Elastic Optical Networks (EONs) and networking virtualization. However, there are many challenges coming along with them, such as Routing and Spectrum Assignment (RSA) in EONs and Virtual Network Embedding (VNE) in network virtualization. This dissertation addresses the algorithm design and analysis for these challenging problems: the impacts of traffic distribution and network topology on lightpath routing, the distance spectrum assignment and the VNE problem for paths and cycles.For lightpath routing, the first subproblem of the RSA, there is always a pending issue that how the changes of the traffic distribution and EON topology affect it. As the lightpath routing plays a critical role in the overall performance of the RSA, this dissertation provides a thoroughly theoretical analysis on the impacts of the aforementioned two key factors. To this end, we propose two theoretical chains, and derive the optimal routing scheme taking into account two key factors. We then treat the second subproblem of RSA, namely spectrum assignment. Any two lightpaths sharing common fiber links might have to be isolated in the spectrum domain with a proper guard-band to prevent crosstalk and/or reduce physical-layer security threats. We consider the scenario with diverse guard-band sizes, and investigate how to assign the spectrum resources efficiently in such a situation. We provide the upper and lower bounds for the optimal solution of the DSA, and further devise an efficient algorithm which can guarantee approximation ratios in some graph classes.The topology heterogeneity of Virtual Network Requests (VNRs) is one important factor hampering the performance of the VNE. However, in many specialized applications, the VNRs are of some common structural features e.g., paths and cycles. To achieve better outcomes, it is thus critical to design dedicated algorithms for these applications by accounting for topology characteristics. We prove the NP-Harness of path and cycle embeddings. To solve them, we propose some efficient algorithms and analyze their approximation ratios.

This phenomenological study presents a map of spatial forces in Ivan Doig’s The Sea Runners. The investigation calls attention to forms of space-experience that come across as a sense of embeddedness in environment. Events, places, feelings, and moods materialize as being nested within greater events and places that are likewise nested in even larger ones. The study shows that experience, place, memory, hope, and narrative have nested structures. The embedding of narrated realities within larger realities is identified as a mode of organization central to the text’s complexity. Even the smallest acts, events, moods, and feelings are set within larger ones with greater scope, reach, or extension. The literary force of The Sea Runners is made possible by a sustained presentation of complexly interlocking orders of embedding. These orders are co-ordinated and synchronized in terms of movement. The study shows how kinetic systems of circulation, vanishing, encircling, and transformation overlap and reinforce each other so as to create a comprehensive co-ordination effect that colours the presentation of landscape and travel. Movement is highlighted in the essay as a factor that makes it possible for these kinetic structures to be fused in various patterns of co-ordination. In The Sea Runners, place and motion complexly combine to shape the narrated flow of lived experience. In its various orders of fluctuation, space-experience flows in intimate association with life-feeling and movement-sensation. Certain basic kinetic categories are delineated as being at the heart of the text’s overall structure. The study brings its findings to a conclusion by discussing these kinetic categories of lived space as running parallel to categories of lived temporality.

This work is dealing with almost-compact embeddings of function spaces, in particular, the class of classical and weak Lorentz spaces with a norm given by a general weight fuction is studied. These spaces are not Banach function spaces in general, thus the almost-compact em- bedding is defined for more general sturctures of rearrangement-invariant lattices. A general characterization of when an r.i. lattice is almost-compactly embedded into a Lorentz space, involving an optimal constant of a certain continuous embedding, is proved. Based on this the- orem and appropriate known results about continuous embeddings, explicit characterizations of mutual almost-compact embeddings of all subtypes of Lorentz spaces are obtained. 1

The goal of this thesis is to investigate the existing graph embedding methods. We aim to represent the nodes of undirected weighted graphs as low-dimensional vectors, also called embeddings, in order to create a rep- resentation suitable for various analytical tasks such as link prediction and clustering. We first introduce several contemporary approaches allowing to create such network embeddings. We then propose a set of modifications and improvements and assess the performance of the enhanced models. Finally, we present a set of evaluation metrics and use them to experimentally evalu- ate and compare the presented techniques on a series of tasks such as graph visualisation and graph reconstruction. 1

Submitted by Caroline Xavier (caroline.xavier@pucrs.br) on 2017-06-08T18:15:48Z No. of bitstreams: 1 TES_RASHA_HASAN_COMPLETO.pdf: 5197991 bytes, checksum: bfd855e20678be1fe11d1731c4cbc317 (MD5) / Made available in DSpace on 2017-06-08T18:15:48Z (GMT). No. of bitstreams: 1 TES_RASHA_HASAN_COMPLETO.pdf: 5197991 bytes, checksum: bfd855e20678be1fe11d1731c4cbc317 (MD5) Previous issue date: 2017-01-16 / Nas ?ltimas tr?s d?cadas de pesquisa em Sistemas Distribu?dos (SDs), um aspecto central discutido ? o de sincronia. Com um sistema ass?ncrono, n?o fazemos suposi??es sobre velocidades de execu??o de processos e / ou atrasos de entrega de mensagens; Com um sistema s?ncrono, fazemos suposi??es sobre esses par?metros [Sch93b]. Sincronismo em SDs impacta diretamente a complexidade e funcionalidade de algoritmos tolerantes a falhas. Uma infra-estrutura s?ncrona contribui para o desenvolvimento de sistemas mais simples e fi?veis, mas tal infra-estrutura ? muito cara e ?s vezes nem sequer vi?vel de implementar. Uma infra-estrutura totalmente ass?ncrona ? mais realista, mas alguns problemas foram mostrados como insol?veis em tal ambiente atrav?s do resultado de impossibilidade por Fischer, Lynch e Paterson [FLP85]. As limita?es tanto em ambientes totalmente s?ncronos como totalmente ass?ncronos levaram ao desenvolvimento de sistemas distribu?dos como s?ncronia parcial [CF99, Ver06]. Em um estudo de funcionalidade de sistemas distribu?dos s?ncronos parciais e de propriedades de Redes Virtuais (RVs), descobrimos que existem v?rios desafios para este tipo de sistemas que podem ser resolvidos com RVs devido ?s propriedades que a virtualiza??o traz. Por exemplo a) partilha de recursos fornecida por RVs permite diminuir o custo ao partilhar a parte s?ncrona da infra-estrutura f?sica, b) isolamento fornecido por a natureza da RVs, isso pode beneficiar os SDs coexistentes na mesma infra-estrutura f?sica que exigem certo n?vel de isolamento,c) resili?ncia garantido atrav?s do processo de aloca??o de recursos de Redes Virtuais, isso permite alocar recursos de reposi??o ao lado dos prim?rios para redes virtuais que exigem garantias de disponibilidade, por exemplo, SDs tolerantes a falhas. Em nosso trabalho, argumentamos que as RVs e um adequado processo de aloca??o de recursos das RVs oferecem um ambiente adequado para executar aplicativos distribu?dos com sincronia parcial. Isto levou ? abstra??o de um novo tipo de RVs: As Redes Virtuais com sincronia h?brida (RVSHs). Nesta tese, apresentamos a id?ia geral das Redes Virtuais com sincronia h?brida motivado pelos SDs com s?ncronia h?brida, e dividimos nosso trabalho em duas partes: a) Espa?oRVSHs propostos pelo SDs com sincronia h?brida em espa?o, e b) Tempo-RVSHs propostos pelo SDs com sincronia h?brida em tempo. No SDs com s?ncronia h?brida em espa?o, a infraestrutura ? composta de subconjuntos de componentes s?ncronos e ass?ncronos, e cada um desses subconjuntos mant?m seu status de sincronia atrav?s do tempo (i.e., os subconjuntos s?ncronos permanecem s?ncronos e os ass?ncronos permanecem ass?ncronos). No SDs com s?ncronia h?brida em tempo, a infra-estrutura ? composta de subconjuntos de n?s e la?os que podem alternar seu status de sincronia atrav?s do tempo (i.e., os componentes se comportam de forma s?ncrona durante os intervalos de tempo e de forma ass?ncrona durante outros intervalos de tempo). As principais contribui??es desta tese s?o: a) caracterizam os RVSHs em seus dois tipos Espa?o-RVSHs e Tempo-RVSHs para refletir tanto a natureza de sincronia em espa?o e em tempo; b) propor uma estrutura adequada para o processo de aloca??o de recursos para ambos Espa?o-RVSHs e Tempo-RVSHs, e c) fornecer uma avalia??o dos modelos propostos para RVSHs. / In the last three decades of research in Distributed Systems (DSs), one core aspect discussed is the one of synchrony. \Vith an asynchronous system, we make no assumptions about process execution speeds andj or message delivery delays; with a synchronous system, we do make assumptions about these parameters [Sch93b]. Synchrony in DSs impacts directly the complexity and functionality of fault-tolerant algorithms. Although a synchronous infrastructure contributes towards the development of simpler and reliable systems, yet such an infrastructure is too expensive and sometimes even not feasible to implemento On the other hand, a fully asynchronous infrastructure is more realistic, but some problems were shown to be unsolvable in such an environment through the impossibility result by Fischer, Lynch and Paterson [FLP85]. The limitations in both fully synchronous or fully asynchronous environments have led to the development of partial synchronous distributed systems [CF99, Ver06]. In a study of partial synchronous distributed systems functionality, and of Virtual Networks (VNs) properties, we found that there are several challenges for this kind of systems that can be solved with VNs due to the properties that virtualization brings. For example a) resources sharing provided by VNs allows decreasing the cost when sharing the synchronous portion of the physical infrastructure, b) isolation provided by the VNs nature can benefit the coexistent DSs on same physical infrastructure that demand certain leveI of isolation, c) resilience guaranteed through the Virtual Networks Embedding (VNE) process that allows allocating spare resources beside the primary ones for virtual networks that require availability guarantees, for example fault tolerant DSs. In our work, we argue that VNs and a suitable VN embedding process offer suitable environment for running distributed applications with partial synchrony. This has led to the abstraction of new type of VNs: The Hybrid Synchrony Virtual Networks (HSVNs). In this thesis, we introduce the general idea of Hybrid Synchrony Virtual Networks (HSVNs) motivated by the hybrid synchronous DSs, and we branch our work into two branches: a) Space-HSVNs addressed to spatial hybrid synchronous DSs, and b) TimeHSVNs addressed to the time hybrid synchronous DSs. In spatial hybrid synchronous DSs, the hybrid synchronous physical infrastructure is composed of subsets of synchronous and asynchronous components, and each of these subsets maintains its synchrony status through time (i.e., synchronous subsets remain synchronous and asynchronous ones remain asynchronous). In time hybrid synchronous DSs, the hybrid synchronous physical infrastructure is composed of subsets of nodes and links that can alternate their synchrony status through time (i.e., the components behave synchronously during time intervals, and asynchronously during other time intervals). The main contributions of this thesis are: a) characterize the HSVNs in its two types Space-HSVNs and Time-HSVNs to reflect both the synchrony space-variant and time-variant nature ofDSs; b) propose a suitable embedding framework for both Space-HSVNs and TimeHSVNs, and c) provide an evaluation of the embedding mo deIs addressed to the HSVNs.

Estudamos inicialmente resultados de classificação de difeomorfismos de produtos de esferas de mesma dimensão. Tratado isto, estudamos os mergulhos suaves de produtos de três esferas, sendo a primeira de dimensão um e as demais de dimensão maior ou igual a um, com a dimensão da última maior ou igual a da segunda, em uma esfera em codimensão um, e buscamos a total caracterização do fecho das duas componentes conexas do complementar de tais mergulhos. Tratamos com enfoque especial os mergulhos do produto de três esferas de dimensão um na esfera de dimensão quatro, e, finalmente, estudamos problemas de classificação de mergulhos PL localmente não-enodados de somas conexas de toros em codimensão um. / We study initially results of classification of difeomorfisms of Cartesian products of spheres of same dimension. Treated this, we study the smooth embeddings of cartesian products of three spheres, being the first one of dimension one and excessively of bigger or equal dimension to one, with the dimension of the last equal greater or of second, in a sphere in codimension one, and search the total characterization of the latch of the two connected components of complementing of such embeddings. We deal with special approach the embeddings of the product to three spheres to dimension one in the sphere dimension four, and, finally, we study problems of classification of PL locally unknotted embeddings of connected sums of torus on codimension one.

Este trabalho apresenta o projeto de um bloco LNA-Misturador dentro de um mesmo circuito integrado para aplicações em um receptor Bluetooth 2;45GHz. Uma estratégia de projeto bem clara, concisa e com uma boa base física e matemática foi desenvolvida para auxiliar o processo de projeto de um bloco LNA-Misturador, composto por um LNA cascode em cascata com um misturador de chaveamento de corrente com entradas simples e degeneração indutiva nas fontes dos estágios de transcondutância. Esta estratégia foi adaptada de trabalhos apresentados na literatura. A estratégia de projeto proposta considera o compromisso entre ruído, linearidade, ganho, dissipação de potência, casamento de impedâncias e isolamento de portas, usando as dimensões dos dispositivos e condições de polarização como variáveis de projeto. Com base nesta estratégia se obteve um bloco LNA-Misturador que atinge algumas especificações propostas. Um bloco LNA-Misturador foi projetado e fabricado em uma tecnologia CMOS 0;35µm para validar a estratégia de projeto proposta. Além disso, para atingir os objetivos, durante o desenvolvimento deste trabalho foi dada atenção especial no projeto dos indutores. Foi projetado, fabricado e medido um chip de teste. Para tal fim foram aplicadas técnicas e estruturas de de-embedding nas medidas para conseguir resultados mais confiáveis. Os resultados experimentais obtidos para os indutores e os resultados preliminares do bloco LNA-Misturador s~ao satisfatórios de acordo com as especificações e os esperados das simulações. No entanto, os indutores integrados degradam significativamente o desempenho do bloco LNA-Misturador. Se forem usados processos de fabricação nos quais os indutores apresentem melhor desempenho, os resultados do bloco LNA-Misturador aplicando a estratégia de projeto desenvolvida neste trabalho podem ser melhorados. Finalmente, é importante ressaltar que a estratégia de projeto proposta neste trabalho já está sendo usada e adaptada em outros projetos com o propósito de melhorar os resultados obtidos, e conseguir auxiliar o processo de projeto deste tipo de blocos. / This work presents a fully integrated LNA-Mixer design for a Bluetooth receiver application at 2:45GHz. A concise design strategy with good physics and mathematics basis was developed to assist the design process of a LNA-Mixer block, formed by a cascode LNA in cascade to a single balanced current commutation Mixer with inductive degeneration. This strategy was adapted from literature and considers the trade-offs between noise, linearity, gain, power dissipation, impedance matching and ports isolation, using the device dimensions and bias conditions as design variables. Based on this strategy, the proposed LNA-Mixer design specifications were achieved. To validate the proposed design strategy, the LNA-Mixer were fabricated in a 0:35µm CMOS process. Furthermore, to achieve the specifications, during the development of this work a special attention to the RF CMOS inductors was given. A test chip was designed, fabricated and measured applying de-embedding structures to obtain more reliable results. The experimental results obtained for the inductors and the preliminary results for the LNA-Mixer are satisfactory compared to the specifications and as expected from simulations. However, the integrated inductors degrade the performance of the block significantly and if a manufacturing process in which the inductor has better performance is used, the resulting LNA-Mixer design applying the strategy developed in this work can be improved. Finally, it is important to highlight that the design strategy proposed in this work is already being used and adapted in other designs in order to improve the results, and to assist the design process of such blocks.

Knowledge bases are employed in a variety of applications, from natural language processing to semantic web search; alas, in practice, their usefulness is hurt by their incompleteness. To address this issue, several techniques aim at performing knowledge base completion, of which embedding models are efficient, attain state-of-the-art accuracy, and eliminate the need for feature engineering. However, embedding models predictions are notoriously hard to interpret. In this work, we propose model-agnostic methods that allow one to interpret embedding models by extracting weighted Horn rules from them. More specifically, we show how the so-called \"pedagogical techniques\", from the literature on neural networks, can be adapted to take into account the large-scale relational aspects of knowledge bases, and show experimentally their strengths and weaknesses. / Bases de conhecimento apresentam diversas aplicações, desde processamento de linguagem natural a pesquisa semântica da web; contudo, na prática, sua utilidade é prejudicada por não serem totalmente completas. Para solucionar esse problema, diversas técnicas focam em completar bases de conhecimento, das quais modelos de embedding são eficientes, atingem estado da arte em acurácia, e eliminam a necessidade de fazer-se engenharia de características dos dados de entrada. Entretanto, as predições dos modelos de embedding são notoriamente difíceis de serem interpretadas. Neste trabalho, propomos métodos agnósticos a modelo que permitem interpretar modelos de embedding através da extração de regras Horn ponderadas por pesos dos mesmos. Mais espeficicamente, mostramos como os chamados \"métodos pedagógicos\", da literatura de redes neurais, podem ser adaptados para lidar com os aspectos relacionais e de larga escala de bases de conhecimento, e mostramos experimentalmente seus pontos fortes e fracos.

Knowledge bases are employed in a variety of applications, from natural language processing to semantic web search; alas, in practice, their usefulness is hurt by their incompleteness. To address this issue, several techniques aim at performing knowledge base completion, of which embedding models are efficient, attain state-of-the-art accuracy, and eliminate the need for feature engineering. However, embedding models predictions are notoriously hard to interpret. In this work, we propose model-agnostic methods that allow one to interpret embedding models by extracting weighted Horn rules from them. More specifically, we show how the so-called \"pedagogical techniques\", from the literature on neural networks, can be adapted to take into account the large-scale relational aspects of knowledge bases, and show experimentally their strengths and weaknesses. / Bases de conhecimento apresentam diversas aplicações, desde processamento de linguagem natural a pesquisa semântica da web; contudo, na prática, sua utilidade é prejudicada por não serem totalmente completas. Para solucionar esse problema, diversas técnicas focam em completar bases de conhecimento, das quais modelos de embedding são eficientes, atingem estado da arte em acurácia, e eliminam a necessidade de fazer-se engenharia de características dos dados de entrada. Entretanto, as predições dos modelos de embedding são notoriamente difíceis de serem interpretadas. Neste trabalho, propomos métodos agnósticos a modelo que permitem interpretar modelos de embedding através da extração de regras Horn ponderadas por pesos dos mesmos. Mais espeficicamente, mostramos como os chamados \"métodos pedagógicos\", da literatura de redes neurais, podem ser adaptados para lidar com os aspectos relacionais e de larga escala de bases de conhecimento, e mostramos experimentalmente seus pontos fortes e fracos.