<p>We study the application of reputation as an instigator of beneficial user behavior in selfish routing and when the network users rely on the network coordinator for information about the network. Instead of using tolls or artificial delays, the network coordinator takes advantage of the users' insufficient data, in order to manipulate them through the information he provides. The issue that arises then is what can be the coordinator's gain without compromising by too much on the trust the users put on the information provided, i.e., by maintaining a reputation for (at least some) trustworthiness.</p> <p>Our main contribution is the modeling of such a system as a repeated game of incomplete information in the case of single-commodity general networks. This allows us to apply known folk-like theorems to get bounds on the price of anarchy that are better than the well-known bounds without information manipulation.</p> / Master of Computer Science (MCS)

Data sets with irrelevant and redundant features and large fraction of missing values are common in the real life application. Learning such data usually requires some preprocess such as selecting informative features and imputing missing values based on observed data. These processes can provide more accurate and more efficient prediction as well as better understanding of the data distribution. In my dissertation I will describe my work in both of these aspects and also my following up work on feature selection in incomplete dataset without imputing missing values. In the last part of my dissertation, I will present my current work on more challenging situation where high-dimensional data is time-involving. The first two parts of my dissertation consist of my methods that focus on handling such data in a straightforward way: imputing missing values first, and then applying traditional feature selection method to select informative features. We proposed two novel methods, one for imputing missing values and the other one for selecting informative features. We proposed a new method that imputes the missing attributes by exploiting temporal correlation of attributes, correlations among multiple attributes collected at the same time and space, and spatial correlations among attributes from multiple sources. The proposed feature selection method aims to find a minimum subset of the most informative variables for classification/regression by efficiently approximating the Markov Blanket which is a set of variables that can shield a certain variable from the target. I present, in the third part, how to perform feature selection in incomplete high-dimensional data without imputation, since imputation methods only work well when data is missing completely at random, when fraction of missing values is small, or when there is prior knowledge about the data distribution. We define the objective function of the uncertainty margin-based feature selection method to maximize each instance's uncertainty margin in its own relevant subspace. In optimization, we take into account the uncertainty of each instance due to the missing values. The experimental results on synthetic and 6 benchmark data sets with few missing values (less than 25%) provide evidence that our method can select the same accurate features as the alternative methods which apply an imputation method first. However, when there is a large fraction of missing values (more than 25%) in data, our feature selection method outperforms the alternatives, which impute missing values first. In the fourth part, I introduce my method handling more challenging situation where the high-dimensional data varies in time. Existing way to handle such data is to flatten temporal data into single static data matrix, and then applying traditional feature selection method. In order to keep the dynamics in the time series data, our method avoid flattening the data in advance. We propose a way to measure the distance between multivariate temporal data from two instances. Based on this distance, we define the new objective function based on the temporal margin of each data instance. A fixed-point gradient descent method is proposed to solve the formulated objective function to learn the optimal feature weights. The experimental results on real temporal microarray data provide evidence that the proposed method can identify more informative features than the alternatives that flatten the temporal data in advance. / Computer and Information Science

The poverty rate among veterans in US has increased over the past decade, according to the U.S. Department of Veterans Affairs (2015). Thus, it is crucial to veterans who live below the poverty level to get sufficient benefit grants. A study on prudently managing health benefit grants for veterans may be helpful for government and policy-makers making appropriate decisions and investments. The purpose of this research is to find an underlying group structure for the veterans' benefit grants dataset and then estimate veterans' benefit grants sought using incomplete data. The generalized linear mixed cluster-weighted model based on mixture models is carried out by grouping similar observations to the same cluster. Finally, the estimates of veterans' benefit grants sought will provide reference for future public policies. / Thesis / Master of Science (MSc)

Design optimality for treatment comparison experiments has been intensively studied by numerous researchers, employing a variety of statistically sound criteria. Their general formulation is based on the idea that optimality functions of the treatment information matrix are invariant to treatment permutation. This implies equal interest in all treatments. In practice, however, there are many experiments where not all treatments are equally important. When selecting a design for such an experiment, it would be better to weight the information gathered on different treatments according to their relative importance and/or interest. This dissertation develops a general theory of weighted design optimality, with special attention to the block design problem. Among others, this study develops and justifies weighted versions of the popular A, E and MV optimality criteria. These are based on the weighted information matrix, also introduced here. Sufficient conditions are derived for block designs to be weighted A, E and MV-optimal for situations where treatments fall into two groups according to two distinct levels of interest, these being important special cases of the "2-weight optimality" problem. Particularly, optimal designs are developed for experiments where one of the treatments is a control. The concept of efficiency balance is also studied in this dissertation. One view of efficiency balance and its generalizations is that unequal treatment replications are chosen to reflect unequal treatment interest. It is revealed that efficiency balance is closely related to the weighted-E approach to design selection. Functions of the canonical efficiency factors may be interpreted as weighted optimality criteria for comparison of designs with the same replication numbers. / Ph. D.

This dissertation consists of three essays that adopt both theoretical and empirical methods of analysis to study certain economies in which the incomplete information and the strategic complementarities between players are important. Chapter 1 explains the topics discussed in the subsequent chapters and gives a brief survey on the literature. In Chapter 2, I revise a traditional global game model by dividing the continuum of players into a group of speculators and a group of stakeholders. It is found that the uniqueness property remains in the new game. Then I extend the static game to a two-stage game and investigate the efficacies of certain label changing mechanisms proposed by the authority to stabilize the regime in the dynamic context. It is shown that a label changing mechanism allowing for downward social mobility may not work, whereas a label changing mechanism allowing for upward social mobility generally makes the regime more stable. In Chapter 3, I add a speculator and an authority to a bank-run model to investigate how the speculator endangers a business or an economy, and what the authority can do about it. In particular, I show that the speculator can increase the financial system's vulnerability by serving as a coordinating device for the investors and thus triggering the crisis. It is further shown that deterring the speculator may not undo the speculator's impact because of multiplicity problem; rewarding holding investors is useless; and eliminating the preemption motives among investors works given enough effort. A discussion of the 1997 Asian financial crisis and the IMF's role in it is also included. Chapter 4 develops a repeated beauty-contest game to investigate the effect of previous winners' actions on the spread of subsequent players' actions. I first characterize the unique equilibrium of the game. Then I focus on the equilibrium dynamics of several variances depicting different forms of action variability. It is found that whether or not a specific variance diminishes over time depends on the relative precision of public and private signals. To illustrate the theoretical results, I conduct an empirical study on the Miss Korea contest. It is found that the contestants' faces have been converging to the ``true beauty'' overall, but diverging from each other over the last 20 years. Chapter 5 concludes. / Ph. D.

Body weight supported treadmill training (BWSTT) is being promoted as an effective means of restoring ambulatory abilities among individuals with incomplete spinal cord injuries. The emphasis of this thesis is on the description of the metabolic demands and the EMG profiles of able-bodied persons and individuals with a spinal cord injury (SCI) while walking under the identical conditions on a body weight support (BWS) treadmill. The secondary purpose was to contrast the metabolic and muscular responses between the two groups. Two separate chapters describing the metabolic demands and EMG profiles respectively follow the review of the literature. The metabolic results indicate that raising the speed and/or decreasing the amount of BWS increase the intensity of BWS treadmill walking, with speed having a more profound effect. The SCI group was less efficient and they had greater metabolic rates of oxygen consumption than the controls for all conditions examined. This led to the conclusion that walking on the treadmill, for the SCI group can provide an effective aerobic exercise stimulus. The EMG profiles suggest that speed and BWS affect the phasic characteristics of the muscular activity while walking for both groups. Furthermore, abnormalities, omissions and inappropriate levels of activity were observed in the SCI group when compared to the controls. These irregularities suggest that the SCI participants have adopted altered motor strategies while walking, relative to the control group. Nonetheless, the SCI participants showed evidence of appropriate modulations in their EMG activity to meet the demands of the task as they changed from one condition to the next. / Thesis / Master of Science (MSc)

本文嘗試利用兩國動態一般模型 (dynamic stochastic general equilibrium, DSGE)架構，分析資本市場存在的不對稱摩擦現象，藉由設定不同兩國債券調整成本的三種形式，模擬兩國總體變數在本國生產力衝擊下的影響分析。本文模型架構主要遵循Bergin et al. (2007)，文中廠商的商品訂價方式採生產者貨幣訂價 (producer currency pricing, PCP)，即廠商的訂價行為依照本國貨幣作為計價單位，透過名目匯率轉換為外國消費者面對的商品價格，單一價格法則 (the law of one price, LOP)在此訂價方式下成立；本文模型的商品訂價方式改採當地貨幣訂價 (local currency pricing, LCP)，本國廠商以當地貨幣作為計價單位訂定其商品價格，該訂價方式廣為近來文獻所採用，並符合已開發國家之訂價行為，故採用此訂價方式。 電腦模擬本文資本市場的三種不同設定，在本國生產力具有 的外生衝擊下，資本邊際生產力的提高致使本國廠商增加投資，代表性個人資金借貸的管道可透過本國債券與外國債券進行融通，而本國代表性個人在買賣本國債券時無需負擔債券調整成本，在購買外國債券時則需負擔債券調整成本，此一設定即在反映資金借貸的摩擦情形；變數的動態模擬結果顯示，資本市場的不對稱摩擦，將透過兩國間的經常帳變化條件，影響資金借貸的流通，對於兩國總體經濟變數具有顯著影響。

This doctoral thesis consists of five research articles on the general topic of optimal decision making under uncertainty in a Bayesian framework. The papers are preceded by three introductory chapters. Papers I and II are dedicated to the problem of finding an optimal stopping strategy to liquidate an asset with unknown drift. In Paper I, the price is modelled by the classical Black-Scholes model with unknown drift. The first passage time of the posterior mean below a monotone boundary is shown to be optimal. The boundary is characterised as the unique solution to a nonlinear integral equation. Paper II solves the same optimal liquidation problem, but in a more general model with stochastic regime-switching volatility. An optimal liquidation strategy and various structural properties of the problem are determined. In Paper III, the problem of sequentially testing the sign of the drift of an arithmetic Brownian motion with the 0-1 loss function and a constant cost of observation per unit of time is studied from a Bayesian perspective. Optimal decision strategies for arbitrary prior distributions are determined and investigated. The strategies consist of two monotone stopping boundaries, which we characterise in terms of integral equations. In Paper IV, the problem of stopping a Brownian bridge with an unknown pinning point to maximise the expected value at the stopping time is studied. Besides a few general properties established, structural properties of an optimal strategy are shown to be sensitive to the prior. A general condition for a one-sided optimal stopping region is provided. Paper V deals with the problem of detecting a drift change of a Brownian motion under various extensions of the classical Wiener disorder problem. Monotonicity properties of the solution with respect to various model parameters are studied. Also, effects of a possible misspecification of the underlying model are explored.

MASTER THESIS ABSTRACT TITLE: Trading Strategy in Incomplete Market AUTHOR: Tomáš Bunčák DEPARTMENT: Department of Probability and Mathematical Statistics, Charles University in Prague SUPERVISOR: Andrea Karlová We focus on the problem of finding optimal trading strategies (in a meaning corresponding to hedging of a contingent claim) in the realm of incomplete markets mainly. Although various ways of hedging and pricing of contingent claims are outlined, main subject of our study is the so-called mean-variance hedging (MVH). Sundry techniques used to treat this problem can be categorized into two approaches, namely a projection approach (PA) and a stochastic control approach (SCA). We review the methodologies used within PA in diversely general market models. In our research concerning SCA, we examine the possibility of using the methods of optimal stochastic control in MVH, and we study the problem of our interest in several settings of market models; involving cases of pure diffusion models and a jump- diffusion case. In order to reach an exemplary comparison, we provide solutions of the MVH problem in the setting of the Heston model via techniques of both of the approaches. Some parts of the thesis are accompanied with numerical illustrations.

Rapid advances in high-throughput data acquisition technologies, such as microarraysand next-generation sequencing, have enabled the scientists to interrogate the expression levels of tens of thousands of genes simultaneously. However, challenges remain in developingeffective computational methods for analyzing data generated from such platforms. In thisdissertation, we address some of these challenges. We divide our work into two parts. Inthe first part, we present a suite of multivariate approaches for a reliable discovery of geneclusters, often interpreted as pathway components, from molecular profiling data with replicated measurements. We translate our goal into learning an optimal correlation structure from replicated complete and incomplete measurements. In the second part, we focus on thereconstruction of signal transduction mechanisms in the signaling pathway components. Wepropose gene set based approaches for inferring the structure of a signaling pathway.First, we present a constrained multivariate Gaussian model, referred to as the informed-case model, for estimating the correlation structure from replicated and complete molecular profiling data. Informed-case model generalizes previously known blind-case modelby accommodating prior knowledge of replication mechanisms. Second, we generalize theblind-case model by designing a two-component mixture model. Our idea is to strike anoptimal balance between a fully constrained correlation structure and an unconstrained one.Third, we develop an Expectation-Maximization algorithm to infer the underlying correlation structure from replicated molecular profiling data with missing (incomplete) measurements.We utilize our correlation estimators for clustering real-world replicated complete and incompletemolecular profiling data sets. The above three components constitute the first partof the dissertation. For the structural inference of signaling pathways, we hypothesize a directed signal pathway structure as an ensemble of overlapping and linear signal transduction events. We then propose two algorithms to reverse engineer the underlying signaling pathway structure using unordered gene sets corresponding to signal transduction events. Throughout we treat gene sets as variables and the associated gene orderings as random.The first algorithm has been developed under the Gibbs sampling framework and the secondalgorithm utilizes the framework of simulated annealing. Finally, we summarize our findingsand discuss possible future directions.