none

São apresentados resultados envolvendo subespaços discretos em diversos tipos de problemas em Topologia Geral. São também apresentadas construções de contraexemplos tanto em ZFC como com axiomas extras. / It is presented some results involving discrete subspaces in many kind of problems in General Topology. It is also presented some constructions of counterexamples in ZFC and assuming extra axioms.

São apresentados resultados envolvendo subespaços discretos em diversos tipos de problemas em Topologia Geral. São também apresentadas construções de contraexemplos tanto em ZFC como com axiomas extras. / It is presented some results involving discrete subspaces in many kind of problems in General Topology. It is also presented some constructions of counterexamples in ZFC and assuming extra axioms.

Lietuvoje sparčiai daugėja nykstančių ežerų, kuriuos būtina valyti. Šiuo metu, jų skaičius jau siekia apie tris tūkstančius. Uždumblėję ežerai daro neigiamą poveikį aplinkai, nyksta ir nebeatlieka savo paskirties. Siekiant sustabdyti ežerų nykimo procesą, didžiausias dėmesys yra skiriamas ežerų valymo būdui naudojant žemsiurbes. Šis valymo būdas yra gana populiarus nors ir brangus. Tačiau ne visos žemsiurbės pritaikytos Lietuvos ežerams valyti. Pagrindinė darbo problema – dumblėjančių Lietuvos ežerų atgaivinimas panaudojant žemsiurbes. Darbo tikslas – įvertinti Lietuvoje naudojamų žemsiurbių tinkamumą ežerams valyti. Darbo uždaviniai: 1. įvertinti gaminamų žemsiurbių tinkamumą Lietuvos ežerų valymui; 2. išanalizuoti šalyje naudojamų žemsiurbių eksploatacinį našumą, pulpos konsistenciją ir transportavimo atstumą; 3. įvertinti iškasamo dumblo tolygumą, valant ežerus; 4. įvertinti Lietuvoje naudojamų žemsiurbių tinkamumą ežerų valymui. Darbo struktūra. Darbą sudaro: įvadas, 3 skyriai, išvados, literatūros sąrašas ir priedai. Darbe pateikti 38 paveikslai, 1 lentelė, 4 priedai. Pirmoje darbo dalyje nagrinėjamas pasirinktos temos aktualumas atliekant literatūros analizę. Antroje darbo dalyje pateikiama tyrimų metodika. Trečioje darbo dalyje pateikiami atlikto tyrimo rezultatai. Darbe aprašomi atlikti trijų Vainiekio, Paežerėlių ir Talkšos ežerų, valytų 2002, 2006, 2007 ir 2008 metais, bei juose naudotų žemsiurbių VT – 2, „Watermaster Classic III“, „Christa Wendland” gauti... [toliau žr. visą tekstą] / The extinction of the lakes, that should be inevitably cleaned, is in great progress in Lithuania. There are about three thousand such kind of lakes. The siltified lakes have the negative influence on the natural environment, they are nearly to extinction and do not serve the purpose. In order to stop the process of the absolute extinction of the lakes, the great attention is paid to the means and methods for cleaning the lakes by using the special machines called suction dredges that are popular, nevertheless, quite expensive. Though, not all kinds of dredges produced are suitable for cleaning the lakes of Lithuania. The main problem of the work- the revival of the siltified lakes by using the suction dredges. The aim of the work - to evaluate the usability of dredges for cleaning the lakes of Lithuania. The objectives of the work: 1. Evaluate the usability of dredges that are produced, for the cleaning of the lakes of Lithuania. 2. Analyze the efficiency of exploitation of the dredges used in Lithuania, the consistence of the solid silt and the distance of transportation of the pulp. 3. Evaluate the even distribution of the silt that is removed. 4. Evaluate the usability of the suction dredges that are exploited in Lithuania. The structure of the work. The parts of the final work are: introduction, 3 sections, conclusions, recommendations, list of literature sources and appendixes. There are 38 pictures, 1 table, 4 appendixes in the final work. In the first part of the... [to full text]

[pt] A presente tese tem por objetivo investigar consistência de desempenho em fundos de ações no Brasil, durante o período ocorrido entre julho/1994 e junho/2001, buscando resposta para a seguinte pergunta: analisando as séries temporais de retornos de fundos de ações no Brasil, será possível determinar aqueles com maior probabilidade de virem a se tornar os de maiores retornos no futuro? Em outras palavras, será o desempenho de fundos de ações consistente ao longo de tempo, permitindo a construção de modelos com capacidade preditiva? A metodologia utilizada foi de caráter exclusivamente quantitativo, hipótese de preços traduzindo informações completas. Foram aplicados testes de consistência de desempenho do tipo Tabelas de Contingência 2X2 (sem ajuste para risco e com ajuste para risco, através de classificação por Alfa s de Jensen) e Portfolio Change Measure (PCM). Os testes PCM utilizam as ponderações das carteiras, sendo tal metodologia de aplicação inédita em fundos de ações no Brasil. Os resultados apresentam evidências de consistência de desempenho entre o primeiro e o período analisados. Tais evidências não se sustentam entre o segundo e o terceiro período. O padrão de consistência é mais forte no fundos de desempenho inferior. Foram aplicados testes para detectar e mensurar impactos de viés de sobrevivência e custos de transação nos retornos dos fundos. Os resultados, obtidos através de ponderação de carteiras e demonstrativos de fontes e aplicações de recursos dos fundos, não detectam viés de sobrevivência nem influência de custos de transações nos retornos dos fundos de ações analisados. O desempenho dos fundos de ações analisados deve ser considerado insatisfatório. / [en] The present dissertation intents to analyze the consistence of performance in stock funds in Brazil, from july/1994 to june/2001. We search the answer to the following question: through the analysis of times series of returns of stock funds in Brazil, is it possible to determine the most likely winners in the future? In other words: the performance of stock funds is consistent along time, allowing for the building of models with predictive power? The methodology applied was a quantitative approach, prices with complete information. I analyzed time series of financial returns of stock funds. The tests for consistence of performance were Contingency Tables 2X2 (with risk adjustment and without risk adjustment through alfa s of Jensen) and Portfolio Change Measure (PCM). The outcomes presents a pattern of consistency between the 1st and the 2nd periods analyzed. This standard is absent between the 2nd and the 3rd periods. The observed pattern of consistence is stronger for the worse performance. We applied tests to detect and measure the effects of survival bias and transaction costs in the returns of funds. The outcomes, obtained trough the analysis of the weighted portfolios and the statements of sources and uses of cash, are the absence of survival bias and effects of transactions costs in the returns of the stock funds analyzed. The performance of the stock funds analyzed must to be considered to be unsatisfactory.

Le contrôle optimal stochastique (en temps discret) s'intéresse aux problèmes de décisions séquentielles sous incertitude. Les applications conduisent à des problèmes d'optimisation degrande taille. En réduisant leur taille, les méthodes de décomposition permettent le calcul numérique des solutions. Nous distinguons ici deux formes de décomposition. La emph{décomposition chaînée}, comme la Programmation Dynamique, résout successivement, des sous-problèmes de petite taille. La décomposition parallèle, comme le Progressive Hedging, consiste à résoudre itérativement et parallèlement les sous-problèmes, coordonnés par un algorithme maître. Dans la première partie de ce manuscrit, Dynamic Programming: Risk and Convexity, nous nous intéressons à la décomposition chaînée, en particulier temporelle, connue sous le nom de Programmation Dynamique. Dans le chapitre 2, nous étendons le cas traditionnel, risque-neutre, de la somme en temps des coûts, à un cadre plus général pour lequel nous établissons des résultats de cohérence temporelle. Dans le chapitre 3, nous étendons le résultat de convergence de l'algorithme SDDP (Stochastic Dual Dynamic Programming Algorithm) au cas où les fonctions de coûts (convexes) ne sont plus polyhédrales. Puis, nous nous tournons vers la décomposition parallèle, en particulier autour des méthodes de décomposition obtenues en dualisant les contraintes (contraintes spatiales presque sûres, ou de non-anticipativité). Dans la seconde partie de ce manuscrit, Duality in Stochastic Optimization, nous commençons par souligner que de telles contraintes peuvent soulever des problèmes de dualité délicats (chapitre 4).Nous établissons un résultat de dualité dans les espaces pairés $Bp{mathrm{L}^infty,mathrm{L}^1}$ au chapitre 5. Finalement, au chapitre 6, nous montrons un résultat de convergence de l'algorithme d'Uzawa dans $mathrm{L}^inftyp{Omega,cF,PP;RR^n}$, qui requière l'existence d'un multiplicateur optimal. La troisième partie de ce manuscrit, Stochastic Spatial Decomposition Methods, est consacrée à l'algorithme connu sous le nom de DADP (Dual Approximate Dynamic Programming Algorithm). Au chapitre 7, nous montrons qu'une suite de problèmes d'optimisation où une contrainte presque sûre est relaxée en une contrainte en espérance conditionnelle épi-converge vers le problème original si la suite des tribus converge vers la tribu globale. Finalement, au chapitre 8, nous présentons l'algorithme DADP, des interprétations, et des résultats de convergence basés sur la seconde partie du manuscrit / Stochastic optimal control addresses sequential decision-making under uncertainty. As applications leads to large-size optimization problems, we count on decomposition methods to tackle their mathematical analysis and their numerical resolution. We distinguish two forms of decomposition. In chained decomposition, like Dynamic Programming, the original problemis solved by means of successive smaller subproblems, solved one after theother. In parallel decomposition, like Progressive Hedging, the original problemis solved by means of parallel smaller subproblems, coordinated and updated by amaster algorithm. In the first part of this manuscript, Dynamic Programming: Risk and Convexity, we focus on chained decomposition; we address the well known time decomposition that constitutes Dynamic Programming with two questions. In Chapter 2, we extend the traditional additive in time and risk neutral setting to more general ones for which we establish time-consistency. In Chapter 3, we prove a convergence result for the Stochastic Dual Dynamic Programming Algorithm in the case where (convex) cost functions are no longer polyhedral. Then, we turn to parallel decomposition, especially decomposition methods obtained by dualizing constraints (spatial or non-anticipative). In the second part of this manuscript, Duality in Stochastic Optimization, we first point out that such constraints lead to delicate duality issues (Chapter 4).We establish a duality result in the pairing $Bp{mathrm{L}^infty,mathrm{L}^1}$ in Chapter 5. Finally, in Chapter 6, we prove the convergence of the Uzawa Algorithm in~$mathrm{L}^inftyp{Omega,cF,PP;RR^n}$.The third part of this manuscript, Stochastic Spatial Decomposition Methods, is devoted to the so-called Dual Approximate Dynamic Programming Algorithm. In Chapter 7, we prove that a sequence of relaxed optimization problems epiconverges to the original one, where almost sure constraints are replaced by weaker conditional expectation ones and that corresponding $sigma$-fields converge. In Chapter 8, we give theoretical foundations and interpretations to the Dual Approximate Dynamic Programming Algorithm

We study the local list number of graphs introduced by Lennerstad and Eriksson. A labelling of a graph on n vertices is a bijection from vertex set to the set {1,…, n}. Given such a labelling c a vertex u is distance consistent if for all vertices v and w |c(u)-c(v)|=|c(u)-c(w)|+1 implies d(u,w)≤ d(u,v). A graph G is k-distance consistent if there is a labelling with k distance-consistent vertices. The local list number of a graph G is the largest k such that G is  k-distance consistent. We determine the local list number of cycles, complete bipartite graphs and some trees as well as prove bounds for some families of trees. We show that the local list number of even cycles is two, and of odd cycles is three. We also show that, if  k, l≥ 3 , the complete bipartite graph  Kk,l has local list number one, the star graph Sn=K1,n has local list number 3, and K2,k  has local list number 2. Finally, we show that for each n≥3 and each k such that 3≤k≤n there is a tree with local list number k. / Vi studerar det lokala listtalet introducerat av Lennerstad och Eriksson. En märkning av en graf på n hörn är en bijektion från hörnmängden till mängden {1, . . . , n}. Givet en sådan märkning c är ett hörn u avståndskonsistent om för alla hörn v och w |c(u) − c(v)| = |c(u) − c(w)| = 1 implicerar d(u, w) ≤d(u, v). En graf G är k-avståndskonsistent om det nns en märkning med k avståndskonsistenta hörn. Det lokala listtalet av en graf G är det största k sådan att G är k -avståndskonsistent. Vi bestämmer den lokala listtalet av cykler, kompletta bipartita grafer och vissa träd och visar som gränser för några familjer av träd. Vi visar att det lokla listtalet av jämna cykler är två, och av udda cykler är tre. Vi visar också att, om k, l ≥ 3, den kompletta bipartita grafen Kk,l har lokalt listtal ett, stjärngrafen Sn = K1,n har lokalt listtal 3, och K2,k har lokalt listtal 2. Slutligen, visar vi att för varje n≥3 och varje k sådant att 3 ≤ k≤n finns ett träd med lokalt listtal k.

Les travaux de cette thèse portent sur le choix du paramètre de lissage dans le problème de l'estimation non paramétrique de la fonction de densité associée à des processus stationnaires ergodiques à temps continus. La précision de cette estimation dépend du choix de ce paramètre. La motivation essentielle est de construire une procédure de sélection automatique de la fenêtre et d'établir des propriétés asymptotiques de cette dernière en considérant un cadre de dépendance des données assez général qui puisse être facilement utilisé en pratique. Cette contribution se compose de trois parties. La première partie est consacrée à l'état de l'art relatif à la problématique qui situe bien notre contribution dans la littérature. Dans la deuxième partie, nous construisons une méthode de sélection automatique du paramètre de lissage liée à l'estimation de la densité par la méthode du noyau. Ce choix issu de la méthode de la validation croisée est asymptotiquement optimal. Dans la troisième partie, nous établissons des propriétés asymptotiques, de la fenêtre issue de la méthode de la validation croisée, données par des résultats de convergence presque sûre. / The work this thesis focuses on the choice of the smoothing parameter in the context of non-parametric estimation of the density function for stationary ergodic continuous time processes. The accuracy of the estimation depends greatly on the choice of this parameter. The main goal of this work is to build an automatic window selection procedure and establish asymptotic properties while considering a general dependency framework that can be easily used in practice. The manuscript is divided into three parts. The first part reviews the literature on the subject, set the state of the art and discusses our contribution in within. In the second part, we design an automatical method for selecting the smoothing parameter when the density is estimated by the Kernel method. This choice stemming from the cross-validation method is asymptotically optimal. In the third part, we establish an asymptotic properties pertaining to consistency with rate for the resulting estimate of the window-width.

[pt] Neste trabalho, investigamos como garantir duas propriedades no desenvolvimento de aplicações multimídia distribuídas interativas: determinismo e consistência. Determinismo é uma propriedade individual dos nós em uma aplicação distribuída e refere-se à característica de um programa sempre produzir a mesma saída a partir de uma mesma entrada. Consistência é uma propriedade de todo o sistema e está relacionada a todos os nós terem sempre a mesma visão da ordem dos eventos. Avaliamos o uso da linguagem síncrona Céu no contexto de programação multimídia para garantir o determinismo. Em relação à consistência, avaliamos se a arquitetura GALS (Globally Asynchronous Local Synchronous) é capaz de prover consistência. Tradicionalmente, aplicações multimídia são desenvolvidas usando linguagens de domínio específico ou linguagens de propósito geral utilizando frameworks especializados. Nenhuma dessas duas abordagens promove o desenvolvimento de aplicações multimídia distribuídas interativas determinísticas e consistentes. Nossa investigação sobre o uso de linguagens síncronas no campo de multimídia levou ao desenvolvimento de Céu-Media, uma biblioteca multimídia determinística para a linguagem síncrona Céu, e Mars, um middleware GALS para aplicações multimídia distribuídas interativas. Os resultados desta tese indicam que, usando as garantias da linguagem síncrona Céu, é possível desenvolver aplicações multimídia determinísticas usando Céu-Media. Além disso, eles também indicam que o modelo de consistência implementado pelo middleware GALS Mars garante que todos os nós sempre concordem com a ordem dos eventos em uma apresentação distribuída. Nós validamos nossa proposta discutindo o desenvolvimento de aplicações multimídia distribuídas propostas pela comunidade de pesquisa usando Céu-Media e Mars, destacando as principais vantagens e também as desvantagens em usar nossa abordagem. / [en] In this, work we investigate how to guarantee two properties in the development of interactive distributed multimedia applications: determinism and consistency. Determinism is a property of individual nodes in a distributed application and states that a program always produces the same output when fed with the same input. Consistency is a property of the whole system and states that all nodes should have the same view of the order of events. We evaluate the use of the synchronous language Céu in the context of multimedia programming for guaranteeing the determinism property. Regarding consistency, we evaluate the GALS (Globally Asynchronous Locally Synchronous) architecture for enforcing consistency. Traditionally, multimedia applications are developed using either a domain specific language or a general purpose language supported by specialized frameworks. Neither of the two approaches promotes the development of deterministic and consistent interactive distributed multimedia applications. Our investigation of the use of synchronous languages in the multimedia field led to the development of Céu-Media, a deterministic multimedia library for the synchronous language Céu, and Mars, a GALS middleware for interactive distributed multimedia applications. The results of this thesis indicate that using the guarantees of the synchronous language Céu it is possible to develop deterministic multimedia applications using Céu-Media. Furthermore, they also indicate that the consistency model enforced by the GALS middleware Mars guarantees that all nodes always agree upon the order of events in a distributed presentation. We validate our proposal by discussing the development of real-world distributed multimedia applications proposed by the research community using both, Céu-Media and Mars, highlighting the main advantages and also the drawbacks of using our approach.

Made available in DSpace on 2015-07-15T13:31:39Z (GMT). No. of bitstreams: 1 Dissertacao de Kionna Oliveira versao PDF.pdf: 1801004 bytes, checksum: 416418b72e9ce91090c13cc2f37adee1 (MD5) Previous issue date: 2006-12-21 / The assessment of the research instrument performance used in the occupational health is of fundamental importance in order to cautiously specify, the trustworthiness of the commonly found situations in work environments. The validation of an instrument has the purpose of verifying whether the mensuration carried out consistently measures what the instrument intends to measure.Nevertheless, to be regarded as a valid one, an instrument must also be reliable,which means that its reliability is required, yet it is not enough for its validity status.This current study of validation has been aimed for assessing the fulfillment of two questionnaires: the Job Content Questionnaire (JCQ), and the Self Reporting Questionnaire (SRQ-20), which were used in the evaluation of labor psychosocial aspects and in the workers mental health respectively. The accomplished research has involved five studies which differed in selected occupations. The analysis was done in two stages; in the first, the evaluation of the SRQ-20 was carried out by examining the ROC curve, factor analysis of tetracoric correlations, multiple correspondence analysis, the internal consistency was evaluated by the formula of Kuder-Richardson (KD-20); in the second stage, the JCQ performance was assessed, in which it has been used statistical procedures such as factor analysis with the extraction method of main components and VARIMAX rotation and calculus of Cronbach s Alpha. The results showed satisfactory indicators of performance for both appraised instruments. The SRQ-20 evaluation revealed the extraction of four factors confirmed by the graphic of the analysis of multiple correspondences, the instrument showed reasonable internal consistency among the scales, the questionnaire performance was satisfactory since it identified symptoms and correctly classified subjects about the suspicion of usual mental disturbs. In general,the JCQ revealed adequacy of its scales to the theoretical model proposed by Karasek, acceptable internal consistency among scales. The good performance of the instruments evaluated in this study showed that these instruments satisfactorily assess what it is intended to be measured and can contribute towards building of indicators of occupational health and efficient policies for the working class. / Scanear