O comportamento individual relativo a viagens sofre a influência de fatores individuais e do meio urbano. Assim, a vizinhança seria uma das variáveis a serem consideradas na análise comportamental relacionada aos deslocamentos. O objetivo principal deste trabalho é analisar a influência da vizinhança no comportamento individual relativo a viagens, através de dados em painel. Dados em painel constituem importante ferramenta em análises comportamentais subjacentes a viagens urbanas, uma vez que propiciam maior quantidade de informações quando comparados aos dados seccionais. Padrões de viagens são mais bem evidenciados, através de dados em painel, caracterizando as habituais rotinas de atividades e viagens, além de melhor identificar comportamentos atípicos. Todavia, a obtenção desses dados comumente não é atividade trivial, demandando recursos monetários e de tempo. Um dos objetivos secundários deste trabalho é apresentar uma maneira prática e pouco onerosa de obtenção de dados em painel através de Smartphones. Tais dados, posteriormente, são aplicados à classificação de indivíduos segundo comportamento relacionado às viagens. A potencialidade da proposta sugerida é validada por meio de um estudo de caso relativo aos estudantes universitários do município de São Carlos - SP, Brasil. Através dos dados em painel, fornecidos pelos estudantes, utilizou-se o algoritmo k-médias considerando quatro variáveis relativas aos deslocamentos. As três categorias obtidas apresentam estrutura espacial e, portanto, possibilitam análises espaciais exploratórias e confirmatórias, almejando a compreensão de influências da vizinhança nas dinâmicas cotidianas. Este trabalho atesta a existência de autocorrelação espacial do conjunto de dados por meio de dois indicadores: Moran e SivarG (Global Spatial Indicator Based on Variogram). A corroboração da dependência espacial, apontada pelos indicadores globais, é confirmada por meio de dois modelos de escolha discreta. Um contendo apenas variáveis originais da base de dados. Outro, análogo ao primeiro, porém adicionado de covariáveis regionais, obtidas por preceitos da geoestatística. A incorporação das covariáveis regionais aumenta a precisão do modelo e promove um incremento das taxas de acertos em validações cruzadas. / Individual travel behaviour is influenced by individual factors and the urban environment. Thus, the neighborhood influence would be one of the variables to be considered in travel behavior analysis related to urban displacements. The main objective of this work is to analyze the influence of neighborhood on travel behavior by panel data. Panel data is an important tool in urban travel behavioral analyzes, since they provide a greater amount of information when compared to sectional data. Travel patterns are more evident through panel data, characterizing the usual routines of activities, as well the atypical behaviors. However, obtaining these data is not a simple task, requiring monetary and time resources. Secondary goals of this work aim to present a practical and inexpensive way to obtain panel data through Smartphones. These data are then applied to the classification of individuals according to travel behavior. The potential of the proposal is validated by a case of study concerning undergraduate and PhD students from São Carlos - SP, Brazil. Using the data provided by the students, a k-means algorithm was used considering four variables regarding displacements. These three categories have spatial structure and allow exploratory and confirmatory spatial data analyzes aiming the comprehension of the nearby influence of data at daily dynamics. This work attests to the existence of spatial autocorrelation of the data set by two indicators: Moran and SivarG (Global Spatial Indicator Based on Variogram). Corroboration of spatial dependence, pointed by the global indicators, is confirmed by two discrete choice models. The first one includes just the original database variables. The second one, analogous to the first, but added of regional covariates obtained by geostatistical concepts. The addition of regional variables leads to a more accurate model, increasing cross-validations hit rates.

O comportamento individual relativo a viagens sofre a influência de fatores individuais e do meio urbano. Assim, a vizinhança seria uma das variáveis a serem consideradas na análise comportamental relacionada aos deslocamentos. O objetivo principal deste trabalho é analisar a influência da vizinhança no comportamento individual relativo a viagens, através de dados em painel. Dados em painel constituem importante ferramenta em análises comportamentais subjacentes a viagens urbanas, uma vez que propiciam maior quantidade de informações quando comparados aos dados seccionais. Padrões de viagens são mais bem evidenciados, através de dados em painel, caracterizando as habituais rotinas de atividades e viagens, além de melhor identificar comportamentos atípicos. Todavia, a obtenção desses dados comumente não é atividade trivial, demandando recursos monetários e de tempo. Um dos objetivos secundários deste trabalho é apresentar uma maneira prática e pouco onerosa de obtenção de dados em painel através de Smartphones. Tais dados, posteriormente, são aplicados à classificação de indivíduos segundo comportamento relacionado às viagens. A potencialidade da proposta sugerida é validada por meio de um estudo de caso relativo aos estudantes universitários do município de São Carlos - SP, Brasil. Através dos dados em painel, fornecidos pelos estudantes, utilizou-se o algoritmo k-médias considerando quatro variáveis relativas aos deslocamentos. As três categorias obtidas apresentam estrutura espacial e, portanto, possibilitam análises espaciais exploratórias e confirmatórias, almejando a compreensão de influências da vizinhança nas dinâmicas cotidianas. Este trabalho atesta a existência de autocorrelação espacial do conjunto de dados por meio de dois indicadores: Moran e SivarG (Global Spatial Indicator Based on Variogram). A corroboração da dependência espacial, apontada pelos indicadores globais, é confirmada por meio de dois modelos de escolha discreta. Um contendo apenas variáveis originais da base de dados. Outro, análogo ao primeiro, porém adicionado de covariáveis regionais, obtidas por preceitos da geoestatística. A incorporação das covariáveis regionais aumenta a precisão do modelo e promove um incremento das taxas de acertos em validações cruzadas. / Individual travel behaviour is influenced by individual factors and the urban environment. Thus, the neighborhood influence would be one of the variables to be considered in travel behavior analysis related to urban displacements. The main objective of this work is to analyze the influence of neighborhood on travel behavior by panel data. Panel data is an important tool in urban travel behavioral analyzes, since they provide a greater amount of information when compared to sectional data. Travel patterns are more evident through panel data, characterizing the usual routines of activities, as well the atypical behaviors. However, obtaining these data is not a simple task, requiring monetary and time resources. Secondary goals of this work aim to present a practical and inexpensive way to obtain panel data through Smartphones. These data are then applied to the classification of individuals according to travel behavior. The potential of the proposal is validated by a case of study concerning undergraduate and PhD students from São Carlos - SP, Brazil. Using the data provided by the students, a k-means algorithm was used considering four variables regarding displacements. These three categories have spatial structure and allow exploratory and confirmatory spatial data analyzes aiming the comprehension of the nearby influence of data at daily dynamics. This work attests to the existence of spatial autocorrelation of the data set by two indicators: Moran and SivarG (Global Spatial Indicator Based on Variogram). Corroboration of spatial dependence, pointed by the global indicators, is confirmed by two discrete choice models. The first one includes just the original database variables. The second one, analogous to the first, but added of regional covariates obtained by geostatistical concepts. The addition of regional variables leads to a more accurate model, increasing cross-validations hit rates.

<p>The Swedish National Debt Office (SNDO) is the Swedish Government’s financial administration. It has several tasks and the main one is to manage the central government’s debt in a way that minimizes the cost with due regard to risk. The debt management problem is to choose currency composition and maturity profile - a problem made difficult because of the many stochastic factors involved. </p><p>The SNDO has created a simulation model to quantitatively analyze different aspects of this problem by evaluating a set of static strategies in a great number of simulated futures. This approach has a number of drawbacks, which might be handled by using a financial optimization approach based on Stochastic Programming. </p><p>The objective of this master’s thesis is thus to apply financial optimization on the Swedish government’s strategic debt management problem, using the SNDO’s simulation model to generate scenarios, and to evaluate this approach against a set of static strategies in fictitious future macroeconomic developments. </p><p>In this report we describe how the SNDO’s simulation model is used along with a clustering algorithm to form future scenarios, which are then used by an optimization model to find an optimal decision regarding the debt management problem. </p><p>Results of the evaluations show that our optimization approach is expected to have a lower average annual real cost, but with somewhat higher risk, than a set of static comparison strategies in a simulated future. These evaluation results are based on a risk preference set by ourselves, since the government has not expressed its risk preference quantitatively. We also conclude that financial optimization is applicable on the government debt management problem, although some work remains before the method can be incorporated into the strategic work of the SNDO.</p>

De nouvelles architectures de traction (hybride, électrique) entrent en concurrence avec les motorisations thermiques conventionnelles. Des batteries Lithium-ion équipent ces véhicules innovants. La durabilité de ces batteries constitue un enjeu majeur mais dépend de nombreux paramètres environ-nementaux externes. Les outils de prédiction de durée de vie actuellement utilisés sont souvent trop simplificateurs dans leur approche. L'objet de ces travaux consiste à caractériser les conditions d'usage de ces batteries (température, tension, courant, SOC et DOD) afin d'étudier avec précision la durée de vie que l'on peut en attendre en fonction de l'application visée. Différents types de véhicules électrifiés (vélos à assistance élec-trique, voitures électriques, voitures hybrides, et trolleybus) ont été instrumentés afin de documenter les conditions d'usage réel des batteries. De larges volumes de données ont été recueillis puis ana-lysés au moyen d'une méthode innovante qui s'appuie sur la classification d'impulsions de courant par l'algorithme des K-means et la génération de cycles synthétiques par modélisation par chaine de Markov. Les cycles synthétiques ainsi obtenus présentent des caractéristiques très proches de l'échantillon complet de données récoltées et permettent donc de représenter fidèlement l'usage réel. Utilisés lors de campagnes de vieillissement de batteries, ils sont susceptibles de permettre l'obtention d'une juste prédiction de la durée de vie des batteries pour l'application considérée. Plusieurs résultats expérimentaux sont présentés afin d'étayer la pertinence de cette approche.

The Swedish National Debt Office (SNDO) is the Swedish Government’s financial administration. It has several tasks and the main one is to manage the central government’s debt in a way that minimizes the cost with due regard to risk. The debt management problem is to choose currency composition and maturity profile - a problem made difficult because of the many stochastic factors involved. The SNDO has created a simulation model to quantitatively analyze different aspects of this problem by evaluating a set of static strategies in a great number of simulated futures. This approach has a number of drawbacks, which might be handled by using a financial optimization approach based on Stochastic Programming. The objective of this master’s thesis is thus to apply financial optimization on the Swedish government’s strategic debt management problem, using the SNDO’s simulation model to generate scenarios, and to evaluate this approach against a set of static strategies in fictitious future macroeconomic developments. In this report we describe how the SNDO’s simulation model is used along with a clustering algorithm to form future scenarios, which are then used by an optimization model to find an optimal decision regarding the debt management problem. Results of the evaluations show that our optimization approach is expected to have a lower average annual real cost, but with somewhat higher risk, than a set of static comparison strategies in a simulated future. These evaluation results are based on a risk preference set by ourselves, since the government has not expressed its risk preference quantitatively. We also conclude that financial optimization is applicable on the government debt management problem, although some work remains before the method can be incorporated into the strategic work of the SNDO.

Mathematical models of signal transduction pathways are characterized by a large number of proteins and uncertain parameters, yet only a limited amount of quantitative data is available. The dissertation addresses this problem using two different approaches: the first approach deals with a model simplification procedure for signaling pathways that reduces the model size but retains the physical interpretation of the remaining states, while the second approach deals with creating rich data sets by computing transcription factor profiles from fluorescent images of green-fluorescent-protein (GFP) reporter cells. For the first approach a model simplification procedure for signaling pathway models is presented. The technique makes use of sensitivity and observability analysis to select the retained proteins for the simplified model. The presented technique is applied to an IL-6 signaling pathway model. It is found that the model size can be significantly reduced and the simplified model is able to adequately predict the dynamics of key proteins of the signaling pathway. An approach for quantitatively determining transcription factor profiles from GFP reporter data is developed as the second major contribution of this work. The procedure analyzes fluorescent images to determine fluorescence intensity profiles using principal component analysis and K-means clustering, and then computes the transcription factor concentration from the fluorescence intensity profiles by solving an inverse problem involving a model describing transcription, translation, and activation of green fluorescent proteins. Activation profiles of the transcription factors NF-κB, nuclear STAT3, and C/EBPβ are obtained using the presented approach. The data for NF-κB is used to develop a model for TNF-α signal transduction while the data for nuclear STAT3 and C/EBPβ is used to verify the simplified IL-6 model. Finally, an approach is developed to compute the distribution of transcription factor profiles among a population of cells. This approach consists of an algorithm for identifying individual fluorescent cells from fluorescent images, and an algorithm to compute the distribution of transcription factor profiles from the fluorescence intensity distribution by solving an inverse problem. The technique is applied to experimental data to derive the distribution of NF-κB concentrations from fluorescent images of a NF-κB GFP reporter system.

Multiple-centered clustering method and decision making system algorithm on menopause data set depending on multiple-centered clustering are described in this study. This method consists of two stages. At the first stage, fuzzy C-means (FCM) clustering algorithm is applied on the data set under consideration with a high number of cluster centers. As the output of FCM, cluster centers and membership function values for each data member is calculated. At the second stage, original cluster centers obtained in the first stage are merged till the new numbers of clusters are reached. Merging process relies upon a &ldquo / similarity measure&rdquo / between clusters defined in the thesis. During the merging process, the cluster center coordinates do not change but the data members in these clusters are merged in a new cluster. As the output of this method, therefore, one obtains clusters which include many cluster centers. In the final part of this study, an application of the clustering algorithms &ndash / including the multiple centered clustering method &ndash / a decision making system is constructed using a special data on menopause treatment. The decisions are based on the clusterings created by the algorithms already discussed in the previous chapters of the thesis. A verification of the decision making system / v decision aid system is done by a team of experts from the Department of Department of Obstetrics and Gynecology of Hacettepe University under the guidance of Prof. Sinan Beksa&ccedil / .

This research is focused on the extraction of visual set patterns in digital images, using relational properties like nearness and similarity measures, as well as descriptive properties such as texture, colour and image gradient directions. The problem considered in this thesis is application of topology in visual set pattern discovery, and consequently pattern generation. A visual set pattern is a collection of motif patterns generated from different unique points called seed motifs in the set. Each motif pattern is a descriptive neighbourhood of a seed motif. Such a neighbourhood is a set of points that are descriptively near a seed motif. A new similarity distance measure based on dot product between image feature vectors was introduced in this research, for image classification with the generated visual set patterns. An application of this approach to pattern generation can be useful in content based image retrieval and image classification.

Cette thèse est consacrée à l'évaluation des familles exponentielles pour les problèmes de la modélisation des bruits et de la segmentation des images couleurs. Dans un premier temps, nous avons développé une nouvelle caractérisation des familles exponentielles naturelles infiniment divisible basée sur la fonction trace de la matrice de variance covariance associée. Au niveau application, cette nouvelle caractérisation a permis de détecter la nature de la loi d'un bruit additif associé à un signal où à une image couleur. Dans un deuxième temps, nous avons proposé un nouveau modèle statistique paramétrique mulltivarié basé sur la loi de Riesz. La loi de ce nouveau modèle est appelée loi de la diagonale modifiée de Riesz. Ensuite, nous avons généralisé ce modèle au cas de mélange fini de lois. Enfin, nous avons introduit un algorithme de segmentation statistique d'image ouleur, à travers l'intégration de la méthode des centres mobiles (K-means) au niveau de l'initialisation pour une meilleure définition des classes de l'image et l'algorithme EM pour l'estimation des différents paramètres de chaque classe qui suit la loi de la diagonale modifiée de la loi de Riesz.

長久以來，信用風險一直是各銀行經營風險中最主要的來源，而就信用風險的衡量部份，巴塞爾委員會希望國際性銀行最低限度必須採用中等複雜程度的風險計算方法。也就是希望銀行能以新巴塞爾協定中信用風險的內部評等法為基本精神建置一套內部自有的信用風險模型來評估交易對手的信用風險。 同時，由於目前國內對於自有信用風險模型的建置和效力驗證的相關研究付之闕如，故本研究以新巴塞爾協定中信用風險的內部評等基礎法為基本精神，並且應用倒傳遞類神經網路方法，建構一套有效的信用風險模型並加以驗證以期能應用於銀行授信決策系統之中，也擬扮演一拋磚引玉的角色，以期未來有更多資源投入相關研究。 首先，本研究藉由文獻探討的方式，決定模型的輸入變數，接著利用ROE來做為評斷企業總體財務表現的指標，同時使用來對上市/櫃公司進行評分，根據評分的結果，再使用K-Means方法來針對所有ROE值為正的上市/櫃公司進行評等等級的切割，以計算所有上市/櫃公司各年度的評等。 研究結果發現： (1) 利用建模資料帶入模型，分別計算每一筆資料的違約機率，也就是該公司當年度的違約機率，再將每一個等級的所有資料的PD值求平均數，即可得到代表該等級的違約機率，而此估計出的違約機率也的確能隨著評等等級的遞增而增加。 因此，當我們要判斷一間公司的違約等級時，可利用本研究所建構出的信用評等模型，估計出該公司違約機率，以判斷該公司的違約等級，以為決策者提供重要的參考依據。 (2) 信用風險預警模型在預測公司下一年度違約與否的能力上，也有不錯的預測準確率；同時，本研究利用預測結果的型I誤差、型II誤差、模型區別率和模型預測率分析來分析預警模型的效度，經實證結果得知，預警模型在效度驗證方面也能有效滿足要求。 由以上的結果得知，本研究所自行發展的信用風險評等模型與信用預警模型相關建構流程、架構與方法論，可有效應用於銀行授信決策系統之中。