由於電腦科技的快速發展，網際網路(World Wide Web；簡稱WWW)使得資料共享及搜尋更為便利，其中的網路搜尋引擎(Search Engine)更是尋找資料的利器，最知名的「Google」公司就是藉由搜尋引擎而發跡。網頁搜尋多半依賴各網頁的特徵，像是熵(Entropy)即是最為常用的特徵指標，藉由使用者選取「關鍵字詞」，找出與使用者最相似的網頁，換言之，找出相似指標函數最高的網頁。藉由相似指標函數分類也常見於生物學及生態學，但多半會計算兩個社群間的相似性，再判定兩個社群是否相似，與搜尋引擎只計算單一社群的想法不同。 本文的目標在於研究若資料服從多項分配，特別是似幾何分配的多項分配（許多生態社群都滿足這個假設），單一社群的指標、兩個社群間的相似指標，何者會有較佳的分類正確性。本文考慮的指標包括單一社群的熵及Simpson指標、兩社群間的熵及相似指標(Yue and Clayton, 2005)、支持向量機(Support Vector Machine)、邏輯斯迴歸等方法，透過電腦模擬及交叉驗證(cross-validation)比較方法的優劣。本文發現單一社群熵指標之表現，在本文的模擬研究有不錯的分類結果，甚至普遍優於支持向量機，但單一社群熵指標分類法的結果並不穩定，為該分類方法之主要缺點。 / Since computer science had changed rapidly, the worldwide web made it much easier to share and receive the information. Search engines would be the ones to help us find the target information conveniently. The famous Google was also founded by the search engine. The searching process is always depends on the characteristics of the web pages, for example, entropy is one of the characteristics index. The target web pages could be found by combining the index with the keywords information given by user. Or in other words, it is to find out the web pages which are the most similar to the user’s demands. In biology and ecology, similarity index function is commonly used for classification problems. But in practice, the pairwise instead of single similarity would be obtained to check if two communities are similar or not. It is dislike the thinking of search engines. This research is to find out which has better classification result between single index and pairwise index for the data which is multinomial distributed, especially distributed like a geometry distribution. This data assumption is often satisfied in ecology area. The following classification methods would be considered into this research: single index including entropy and Simpson index, pairwise index including pairwise entropy and similarity index (Yue and Clayton, 2005), and also support vector machine and logistic regression. Computer simulations and cross validations would also be considered here. In this research, it is found that the single index, entropy, has good classification result than imagine. Sometime using entropy to classify would even better than using support vector machine with raw data. But using entropy to classify is not very robust, it is the one needed to be improved in future.

Um sistema reparável opera sob uma estratégia de manutenção que exige ações de recuperação preventiva em tempos pré-definidos e ações de reparo quando ocorre a perda de função do sistema. A manutenção preventiva (MP) é programada periodicamente e muitas vezes possui um intervalo de tempo fixo para ações. No entanto, as atividades de MP podem não restaurar o sistema para uma condição similar ao início de vida deste, mas para uma situação intermediária. Nesse caso, a MP é denominada de imperfeita. Além disso, ao longo da vida do sistema, são executados diferentes planos de manutenção com condições e atividades distintas que podem afetar a intensidade de falha de diferentes maneiras. Para modelar essas características da MP em um sistema reparável, propõe-se uma nova classe de modelo de fator de melhoria, denominado fator de melhoria variável que possibilita a modelagem da situação de manutenção perfeita. A formulação da função de verossimilhança foi desenvolvida para estimação dos parâmetros bem como desenvolvidos testes de verificação da qualidade de ajuste, intervalos de confiança para os parâmetros e otimização da periodicidade de realização da MP com base no enfoque dos novos modelos propostos. Os resultados foram aplicados em dados reais e verificou-se uma parametrização mais flexível a MP imperfeita e maior versatilidade nas análises de confiabilidade do sistema quando utilizado os novos modelos. / A repairable system operates under a maintenance strategy that calls for preventive repair actions at prescheduled times and the repair actions that restore system when failure occurs. The preventive maintenance (PM) is scheduled periodically and it often holds a fixed time interval for PM actions. However, PM activities are generally imperfect and cannot restore the system to as good as new condition but to an intermediate situation, which is called imperfect PM. In addition, throughout system life are implemented diverse maintenance policies with different activities and conditions that may affect the failure intensity in different ways. To model these PM characteristics, proposes a new model class of improvement factor called variable improvement factor that also enables modeling perfect maintenance situation. The likelihood function is developed for parameter estimation as well as goodness-of-fit tests and confidence intervals for the parameters are developed, and optimization of the PM intervals based on the proposed models is presented. The proposed model was applied to a data set and a more flexible parameterization for imperfect PM and greater versatility in the system reliability analysis were verified with the use of the new model.

Il existe deux façons d'analyser l'activité cérébrale acquise en magnétoencéphalographie (MEG) : soit en moyennant les réponses suscitées par la répétition d'un stimulus afin d'observer le « champ évoqué »; soit en décomposant le signal en bandes oscillatoires (tel que l'alpha, le bêta ou le gamma), chacune étant associée à différents rôles fonctionnels. Ces méthodes ne prennent cependant pas compte de la complexité de l'activité cérébrale dont l'essentiel est arythmique, notamment au repos. Pour pallier à cela, une autre approche consiste à analyser le spectre de puissance en 1/f observable dans les très basses fréquences, une caractéristique des systèmes dont la dynamique est invariante d'échelle. Pour savoir si cette propriété joue un quelconque rôle dans le fonctionnement cérébral et si elle a des conséquences sur le comportement, nous avons établit un paradigme d'apprentissage visuel permettant d'observer de la plasticité fonctionnelle au cours d'une session MEG. Pour avoir un entraînement optimal, nous avons développé de nouveaux stimuli audiovisuels (AV) (une texture acoustique associée à un nuage de points colorés en mouvement) permettant une intégration multisensorielle et de ce fait un meilleur apprentissage que celui apporté par un entraînement visuel seul (V) ou accompagné d'un bruit acoustique (AVn). Nous avons ensuite étudié les corrélats neuronaux de ces trois types d'apprentissage par l'analyse classique des champs évoqués. Une fois l'activité reconstruite sur la surface corticale de chaque individu à l'aide de MNE-dSPM, nous avons identifié le réseau impliqué dans la tâche au sein de chaque groupe. En particulier, la plasticité sélective observée dans l'aire hMT+ associée au traitement du mouvement visuel corrélait avec les progressions comportementales des individus et était soutenue en AV par un plus vaste réseau comprenant notamment des aires multisensorielles. Parallèlement, nous avons exploré les liens reliant le comportement et les propriétés d'invariance d'échelle de ces mêmes signaux MEG reconstruits sur le cortex. Tandis que la plupart des études se limitent à analyser l'auto-similarité (une caractéristique globale synonyme de longue mémoire), nous avons aussi considéré les fluctuations locales (c-à-d la multifractalité) au moyen de l'analyse WLBMF. Nous avons trouvé des modulations couplées de l'auto-similarité et de la multifractalité dans des régions similaires à celles révélées par l'analyse des champs évoqués. Plus surprenant, Le degré de multifractalité relevé dans chaque individu convergeait durant l'entraînement vers un même attracteur reflétant la performance comportementale asymptotique. / The analysis of Human brain activity in magnetoencephalography (MEG) can be generally conducted in two ways: either by focusing on the average response evoked by a stimulus repeated over time, more commonly known as an ``event-related field'' (ERF), or by decomposing the signal into functionally relevant oscillatory or frequency bands (such as alpha, beta or gamma). However, the major part of brain activity is arrhythmic and these approaches fail in describing its complexity, particularly in resting-state. As an alternative, the analysis of the 1/f-type power spectrum observed in the very low frequencies, a hallmark of scale-free dynamics, can overcome these issues. Yet it remains unclear whether this scale-free property is functionally relevant and whether its fluctuations matter for behavior. To address this question, our first concern was to establish a visual learning paradigm that would entail functional plasticity during an MEG session. In order to optimize the training effects, we developed new audiovisual (AV) stimuli (an acoustic texture paired with a colored visual motion) that induced multisensory integration and indeed improved learning compared to visual training solely (V) or accompanied with acoustic noise (AVn). This led us to investigate the neural correlates of these three types of training using first a classical method such as the ERF analysis. After source reconstruction on each individual cortical surface using MNE-dSPM, the network involved in the task was identified at the group-level. The selective plasticity observed in the human motion area (hMT+) correlated across all individuals with the behavioral improvement and was supported by a larger network in AV comprising multisensory areas. On the basis of these findings, we further explored the links between the behavior and scale-free properties of these same source-reconstructed MEG signals. Although most studies restricted their analysis to the global measure of self-similarity (i.e. long-range fluctuations), we also considered local fluctuations (i.e. multifractality) by using the Wavelet Leader Based Multifractal Formalism (WLBMF). We found intertwined modulations of self-similarity and multifractality in the same cortical regions as those revealed by the ERF analysis. Most astonishing, the degree of multifractality observed in each individual converged during the training towards a single attractor that reflected the asymptotic behavioral performance in hMT+. Finally, these findings and their associated methodological issues are compared with the ones that came out from the ERF analysis.

<div>Free surface flows where the shape of the interface separating two fluids is unknown <i>apriori</i> are an important area of interest in fluid dynamics. The study of free surface flows such as the breakup and coalescence of drops, and thinning and rupture of films lends itself to a diverse range of industrial applications, such as inkjet printing, crop spraying, foam and emulsion stability, and nanolithography, and helps develop an understanding of natural phenomena such as sea spray generation in oceans, or the dynamics of tear films in our eyes. In free surface flows, singularities are commonly observed in nite time, such as when the radius of a thread goes to zero upon pinchoff or when the thickness of a film becomes zero upon rupture. Dynamics in the vicinity of singularities usually lack a length scale and exhibit self-similarity. In such cases, universal scaling laws that govern the temporal behavior of measurable physical quantities such as the thickness of a lm can be determined from asymptotic analysis and veried by high-resolution experiments and numerical simulations. These scaling laws provide deep insight into the underlying physics, and help delineate the regions of parameter space in which certain forces are dominant, while others are negligible. While the majority of previous works on singularities in free-surface flows deal with Newtonian fluids, many fluids in daily use and industry exhibit non-Newtonian rheology, such as polymer-laden, emulsion, foam, and suspension flows.</div><div><br></div><div><div>The primary goal of this thesis is to investigate the thinning and rupture of thin films of non-Newtonian fluids exhibiting deformation-rate-thinning (power-law) rheology due to attractive intermolecular van der Waals forces. This is accomplished by means of intermediate asymptotic analysis and numerical simulations which utilize a robust Arbitrary Eulerian-Lagrangian (ALE) method that employs the Galerkin/Finite-Element Method for spatial discretization. For thinning of sheets of power-law fluids, a signicant finding is the discovery of a previously undiscovered scaling regime where capillary, viscous and van der Waals forces due to attraction between the surfaces of the sheet, are in balance. For thinning of supported thin films, the breakdown of the lubrication approximation used almost exclusively in the past to study such systems, is shown to occur for films of power-law fluids through theory and conrmed by two dimensional simulations. The universality of scaling laws determined for rupture of supported films is shown by studying the impact of a bubble immersed in a power-law fluid with a solid wall.</div></div><div><br></div><div><div>Emulsions, which are ne dispersions of drops of one liquid in another immiscible liquid, are commonly encountered in a variety of industries such as food, oil and gas, pharmaceuticals, and chemicals. Stability over a specied time frame is desirable in some applications, such as the shelf life of food products, while rapid separation into its constituent phases is required in others, such as when separating out brine from crude oil. The timescale over which coalescence of two drops of the dispersed phase occurs is crucial in determining emulsion stability. The drainage of a thin film of the outer liquid that forms between the two drops is often the rate limiting step in this process. In this thesis, numerical simulations are used to decode the role played by fluid inertia in causing drop rebound, and the subsequent increase in drainage times, when two drops immersed in a second liquid are brought together due to a compressional flow imposed on the outer liquid. Additionally, the influence of the presence of insoluble surfactants at the drop interface is studied. It is shown that insoluble surfactants cause a dramatic increase in drainage times by two means, by causing drop rebound for small surfactant concentrations, and by partially immobilizing the interface for large surfactant concentrations.</div></div>

O presente trabalho consiste em uma extensão do sistema FreeFlow-2D para simular escoamentos de fluidos não newtonianos bidimensionais com superfí cies livres, onde o fluido é descrito pelos modelos de Cross ou o modelo ``power-law\'\'. O método numérico empregado é o método GENSMAC. As equações governantes são aproximadas pelo método de diferenças finitas em uma malha deslocada e partículas marcadoras são utilizadas para a visualização do escoamento e localização da superfície livre. Resultados numéricos são apresentados. Em particular, a presente implementação é validada comparando-se a solução numérica com uma solução analítica / This work presents an extention of the Freeflow-2D system to non-Newtonian free surface flows. The governing equations are solved by the finite difference method on a staggered grid. Marker particles are used to describe the fluid providing the location and the visualization of the free surface. The methodology employed is based on the GENSMAC method. The fluid is modelled by the Cross and power-law models. Numerical examples are presented. The code is validated by making a comparison between analytical and numerical solutions

Submitted by Silvana Teresinha Dornelles Studzinski (sstudzinski) on 2017-02-08T14:37:06Z No. of bitstreams: 1 Henrique Zenker dos Santos_.pdf: 9480964 bytes, checksum: 5a3a14f9704ce1135e84db57590283a7 (MD5) / Made available in DSpace on 2017-02-08T14:37:06Z (GMT). No. of bitstreams: 1 Henrique Zenker dos Santos_.pdf: 9480964 bytes, checksum: 5a3a14f9704ce1135e84db57590283a7 (MD5) Previous issue date: 2016-10-21 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Países desenvolvidos apresentam crescimento contínuo nas taxas globais de consumo de energia. Um dos fatores preponderantes está relacionado ao uso de condicionadores de ar para manutenção das condições de conforto no interior de edificações, onde tal parcela já atinge valores de consumo superiores aos dos setores de indústria e transporte. Sabe-se que o consumo devido ao condicionamento de ar afeta diretamente o desempenho energético e ambiental de uma edificação e, por conseguinte, a qualidade de habitação e de vida dos usuários. No Brasil este processo ocasiona um vertiginoso incremento nos gastos públicos, com inúmeras consequências de ordem econômica e social. Dados do Ministério de Minas e Energia indicam que 20% da energia consumida no país é destinada ao abastecimento de residências. Cenário este que tende a evoluir devido aos constantes lançamentos do mercado imobiliário, derivativos de planos habitacionais estimulados pelo governo federal (como o Minha Casa Minha Vida), em decorrência do déficit habitacional registrado no país. Atrelado a isso observa-se o baixo padrão construtivo das edificações e o emprego de materiais e esquadrias de qualidade igualmente diminutas, que acarretam em ineficiências de carga térmica. A taxa de infiltração em uma edificação é regida pela relação estabelecida por uma equação de lei de potência, caracterizada pela presença de um coeficiente de escoamento “C” e um expoente de escoamento “n” (geralmente da ordem de 0,667 [-]). Esta dissertação apresenta, para uma esquadria de alumínio, modelo de correr, o coeficiente de escoamento C = 0,028 [dm³/s.m.Pan] e o expoente de escoamento “n” 0,552 [-]. Para uma esquadria de PVC, modelo integrado, o coeficiente de escoamento C = 0,022 [dm³/s.m.Pan] e o expoente de escoamento “n” 0,605 [-]. / Developed countries have continued growth in global energy consumption rates. One of the most important factors is related to the use of air conditioners to the comfort conditions inside buildings’ maintenance, which such a part already reaches consumption values higher than the industry and transportation sectors. It is known that consumption due to air conditioning directly affects the energy and environmental performance of a building and therefore the house’s quality and the people’s life. In Brazil, this process leads to a big increase in public spending, with many consequences of economic and social order. Information from the Ministry of Mines and Energy indicate that 20% of the energy consumed in the country is used for supplying homes. This scenario tends to evolve due to the constant housing market releases, derived from housing plans stimulated by the federal government (such as “Minha Casa Minha vida”), in consequence of the housing deficit registered in the country. Related to this, it is observed the low construction standards of buildings and the use of materials and frames of low quality that lead to thermal load inefficiencies. Engineers and architects spend a part of their activities in creating scenarios for evaluating the thermal load calculation and vulnerability due to the influence caused by air infiltration into the environment. The infiltration rate of a building is managed by the established relationship by a power law equation, which is characterized by a flow coefficient "C" and a flow exponent "n" (usually on the value of 0.667 [-]). This work shows for an aluminum window, sliding template, the flow coefficient C = 0.028 [dm³/s.m.Pan] and the flow exponent "n" 0.552 [-]. For PVC window, integrated model, the flow coefficient C = 0.022 [dm³/s.m.Pan] and the exponent flow "n" 0.605 [-].

Les nanoparticules d’oxyde de fer ont récemment été envisagées comme outils pour l’ingénierie tissulaire. Elles sont internalisées par les cellules qui deviennent alors magnétiques. Des forces magnétiques peuvent ainsi être appliquées à distance sur ces cellules pour contrôler leur organisation spatiale et temporelle, et former un tissu. Ces applications posent la question du devenir des nanoparticules, qui conditionne in fine leur utilisation clinique. Ce travail s’inscrit dans ce cadre et comporte deux axes.La première partie traite de l’étude des propriétés mécaniques et rhéologiques de tissus biologiques modèles, les agrégats multicellulaires. Une combinaison de méthodes magnétiques est proposée pour fabriquer et stimuler des tissus magnétiques de taille et de forme contrôlées. Ces agrégats magnétiques sont soumis à distance à des contraintes magnétiques d’écrasement. L’étude de leur déformation permet d’explorer des caractéristiques statiques et dynamiques rarement étudiées à l’échelle tissulaire (tension de surface, loi puissance, non linéarité). La deuxième partie se concentre sur l'évolution à moyen terme des nanoparticules dans leur environnement tissulaire, au cœur des agrégats. En combinant ce tissu modèle avec des méthodes de quantification magnétique, nous avons pu mettre en évidence une dégradation massive d’origine endosomale, sans pour autant impacter de manière importante l’homéostasie du fer. De plus, le modèle tissulaire mis en place permet d’étudier la biodégradation intracellulaire de n’importe quel type de nanoparticules. Nous l'avons testé avec des nano-architectures plus complexes: nanocubes, nanodimers, ou nanoparticules magnéto-plasmoniques / Iron oxide nanoparticles are promising candidates for applications in nanomedecine (contrast agents, vectors). They were also recently considered as a powerful tool for tissue engineering. Cells, magnetized through nanoparticules internalization, can be organized in space and time thanks to remote magnetic forces. For all those applications the nanoparticles fate inside the cells remains a key issue concerning the final clinical use. The first part of this work focuses on the study of the mechanical and rheological properties of biological tissue models, the multicellular aggregates. An original magnetic molding method and two different experimental setups were developed to produce aggregates with controlled shapes and sizes, to measure their surface tension and to evidence their power law and non linear behavior.In the second part, we investigate the medium term fate of iron oxide nanoparticles in stem cells forming a spheroid as a model tissue. We reveal a massive endosomal degradation. The set of methods and spheroid model we propose allow a comprehensive and quantitative follow up of the biodegradation of any nanomaterials. This was illustrated by investigating the degradation of nanomaterials with more complex nano-architectures (nanocubes, nanodimers) and assessing the efficiency of a protection strategy to modulate the biodegradation

Un suivi des processus microstructuraux prenant place au cours de la fatigue oligocyclique de l’aluminium pur est assuré par la technique d’émission acoustique EA par ces deux types: émission continue et discrète. Cette technique est intéressante car elle permet de suivre l’évolution dynamique de la structure tout le long de l’essai. Les différents stades du comportement macroscopique du matériau au cours des sollicitations cycliques sont clairement différenciés par l’activité acoustique. Nous distinguons cinq stades : écrouissage primaire, adoucissement primaire, écrouissage secondaire, adoucissement secondaire et rupture. Les trois premiers stades mettent en jeu des phénomènes microstructuraux liés à la plasticité du matériau tandis que des phénomènes relatifs à l’endommagement (micro et macro-fissuration) dominent les derniers stades. L’EA continue résulte de l’effet cumulatif de nombreux mouvements de dislocations de faible amplitude et décorrélés entre eux. Cette plasticité continue diminue au cours du 1er stade mais copie l’évolution de la réponse macroscopique de l’échantillon au cours des stades suivants. Ce comportement est lié aux structures de dislocations établies à travers les différents stades de fatigue. En revanche, l’EA de type discret enregistrée lors des trois premiers stades est associée à un autre type de plasticité : la plasticité intermittente, se manifestant à travers des mouvements coopératifs de grande ampleur, les avalanches de dislocations. Ces avalanches de dislocations génèrent des signaux acoustiques de tailles variables, distribuées en loi de puissance. La plasticité intermittente est alors invariante d’échelle tandis que la plasticité continue met en jeu des mouvements ayant une taille caractéristique. Nous mettons ainsi en évidence pour la première fois la coexistence de ces deux types de plasticité dans un matériau cubique à faces centrées CFC, qui ne sont donc pas incompatibles. Au cours des deux derniers stades de fatigue, les signaux acoustiques enregistrés se catégorisent également en deux groupes: l’un est caractérisé par des invariances d’échelle, l’autre associé à une taille caractéristique. La première catégorie comprend des signaux acoustiques indépendants, apparaissant aléatoirement au cours des cycles. Ces signaux sont générés par des phénomènes de microfissuration au sein du volume de l’échantillon (nucléation, percolation…). Le second groupe, réunit des signaux acoustiques générés quasiment au même niveau de contrainte sur plusieurs cycles successifs et ayant une signature acoustique quasi identique. Nous nommons ces signaux multiplets en référence à la sismologie. Nous émettons l'hypothèse que de tels multiplets d’EA sont la signature de la propagation, cycle après cycle, d'une fissure de fatigue dont la trace peut être vu post-mortem avec les stries de fatigue sur une surface de fracture, ou encore la signature de frottements entre les aspérités présentes de part et d’autre des lèvres de fissures. / An analysis of microstructural processes taking place during low-cycle fatigue of pure aluminum is performed by the Acoustic Emission technique (AE) with its two types: continuous and discrete. The main interest of this technique is that it enables the following of the dynamic evolution of the microstructure during the fatigue test. We distinguished five fatigue stages: primary hardening, primary softening, secondary hardening, secondary softening and failure. The various stages of the material’s macroscopic behavior during cyclic loading are clearly differentiated by the acoustic activity. During the first three stages, mainly microstructural phenomena related to plasticity of material are taking place, whereas damage (micro and macro-cracking) dominate the last two stages. The continuous AE results from the cumulative effect of many uncorrelated dislocations’ movements of low amplitude. This continuous plasticity decreases during the 1st stage but reproduces the evolution of the macroscopic behavior of the sample during following stages. This behavior is related to the dislocation structure established during the various fatigue stages. On the other hand, the discrete AE recorded at the time of the first three stages is associated to another type of plasticity: intermittent plasticity. This plasticity is associated to co-operative dislocation movements of great amplitude; dislocation avalanches. These dislocation avalanches generate acoustic signals power law distributed in amplitude and energies. Intermittent plasticity is then scale invariant while continuous plasticity is associated to dislocation movements with a characteristic size. We highlight for the first time the coexistence of these two types of plasticity in FCC materials, which are therefore not incompatible. During the last two stages of fatigue, the recorded acoustic signals are categorized in two groups: the first one is characterized by scale invariance whereas the other is associated to a characteristic size. The first category comprises independent acoustic signals, appearing randomly during cycles. These signals are generated by micro-cracking events within the volume of the sample (nucleation, percolation…). The second group contains acoustic signals generated almost at the same stress level during several successive cycles and having a nearly identical acoustic signature. We name these signals multiplets in reference to seismology. We put forth the hypothesis that such AE multiplets are the signature of fatigue crack propagation, one cycle after the other, whose trace can be observed post-mortem with fatigue striations on fracture surface, or a signature of frictions between the asperities present on both sides of the crack.

預測市場的研究近年來在學界逐漸受到重視，因為它利用價格具有訊息加總的功能，每每創造出良好的預測績效，但一個預測市場的建立在諸多原因下，通常不易吸引大規模的參與者，例如為免觸犯法令規定，以虛擬貨幣代替真錢進行交易，在缺乏真實貨幣的獲利誘因下，很難有效吸引參與者，即便真能以真實貨幣交易，若實驗的議題並非一般大眾感興趣的話題也不易吸引多數人參與，在這種情況下無法避免要面臨市場交易過於淺薄的問題，雖然不少文獻標榜淺薄市場不會影響預測市場的預測精準度，但並不表這是一個可以置之不理的問題。 本文以預測市場預測2006年北高市長選舉為實證基礎，闡明淺薄市場對價格產生的影響，以及這些影響將導致對未來事件的錯誤評價與推論，要避免這種錯誤的評價與推論唯有設法消除淺薄市場引發的干擾，因此我們提出了五種可以消除這些干擾的方法並從中選擇一較佳者。如同一般文獻的讚揚，我們再次從預測市場獲得精確的預測效果，同時證明所謂淺薄市場不影響預測市場的預測精準度前提乃在消除淺薄市場對價格產生的干擾之後才能還原這個真相。

Many complex systems arising from nature and human society can be described as complex networks. In this dissertation, on the basis of complex network theory, we pay attention to the topological structure of complex network and the dynamics on it. We established models to investigate the influences of the structure on the dynamics of networks and to shed light on some peculiar properties of complex systems. This dissertation includes four parts. In the first part, the empirical properties (degree distribution, clustering coefficient, diameter, and characteristic path length) of urban road network of Le Mans city in France are studied. The degree distribution shows a double power-law which we studied in detail. In the second part, we propose two models to investigate the possible mechanisms leading to the deviation from simple power law. In the first model, probabilistic addition of nodes and links, and rewiring of links are considered; in the second one, only random and preferential link growth is included. The simulation results of the modelling are compared with the real data. In the third part,the probabilistic uncertainty behavior of double power law distribution is investigated. The network optimization and optimal design of scale free network to random failures are discussed from the viewpoint of entropy maximization. We defined equilibrium network ensemble as stationary ensembles of graphs by using some thermodynamics like notions such as "energy", "temperature", "free energy" for network. In the forth part, an union-division model is established to investigate the time evolution of certain networks like cultural or economical networks. In this model, the nodes represent, for example, the cultures. Several quantities such as richness, age, identity, ingredient etc. are used to parameterize the probabilistic evolution of the network. The model offers a long term view on the apparently periodic dynamics of an ensemble of cultural or economic entities in interaction.