Investment of mutual fund has always been popular to the common investor. No matter investing at Systematic Investment Plan or One-time Investment, choosing a ¡§good¡¨ mutual fund is expected for every investor. However, how to find a good one among a large number of funds? In the past reports or theses, there are many reasons of influencing mutual fund performance, which include fund¡¦s size among others. The common sense said that ¡§if you want to invest in mutual fund, you should choose the larger size ¡¨ Is that true? Through the past reports we can find out different opinions. This study mainly research into the relationship between performance and size of domestic open-ended mutual funds. Expect to give a simpler data for reference and lower the risk before general investors put money into funds. The logic of Spearman's Rank Correlation coefficient would be applied as the statistical inference in this study, and the domestic open-ended equity fund, balanced fund, and quasi-money market fund will be used as the observed samples. Monthly performance and size is the observed unit of time in this research period from March 2001 to February 2010. Thus via this mode will know whether the correlation between fund¡¦s performance and size is positive, negative or there is no correlation. The conclusions of this study are presented as follows: 1. There is a low positive correlation between performance and size for equity fund and quasi-money market fund. But the balanced fund does not show any correlation for these two variables. 2. Moreover, the study shows there is better performance for investing in over 4 billion NT asset of equity fund in a short term period. But the result of balanced fund and quasi-money market fund can¡¦t highlight the optimal size in a short term period. 3. From the long term point of view, the optimal interval of size for equity and balanced fund is between 1 billion and 2 billion. And the optimal size to quasi-money market fund is over 40 billion.

Most American universities and colleges require students to provide faculty evaluation at end of each academic term, as a way of measuring faculty teaching performance. Although some analysts think that this kind of evaluation does not necessarily provide a good measurement of teaching effectiveness, there is a growing agreement in the academic world about its reliability. This study attempts to find any strong statistical evidence supporting faculty evaluation by students as a measure of faculty teaching effectiveness. Emphasis will be on analyzing relationships between instructor ratings by students and corresponding students’ grades. Various statistical methods are applied to analyze a sample of real data and derive conclusions. Methods considered include multivariate statistical analysis, principal component analysis, Pearson's correlation coefficient, Spearman's and Kendall’s rank correlation coefficients, linear and logistic regression analysis.

The practice of ranking objects, events, and people to determine relevance, importance, or competitive edge is ancient. Recently, the use of rankings has permeated into daily usage, especially in the fields of business and education. When determining the association among those creating the ranks (herein called sources), the traditional assumption is that all sources compare a list of the same items (herein called elements). In the twenty-first century, it is rare that any two sources choose identical elements to rank. Adding to this difficulty, the number of credible sources creating and releasing rankings is increasing. In statistical literature, there is no current methodology that adequately assesses the association among multiple sources. We introduce rank-adapted singular value decomposition (R-A SVD), a new method that uses Kendall's tau as the underlying correlation method. We begin with (P), a matrix of data ranks. The first step is to factor the covariance matrix (K) as follows: K = cov(P) = V D^2 V Here, (V) is an orthonormal basis for the rows that is useful in identifying when sources agree as to the rank order and specifically which sources. D is a diagonal of eigenvalues. By analogy with singular value decomposition (SVD), we define U^* as U^* = PVD^(-1) The diagonal matrix, D, provides the factored eigenvalues in decreasing order. The largest eigenvalue is used to assess the overall association among the sources and is a conservative unbiased method comparable to Kendall's W. Anderson's test determines whether this association is significant and also identifies other significant eigenvalues produced by the covariance matrix.. Using Anderson's test (1963) we identify the a significantly large eigenvalues from D. When one or more eigenvalues is significant, there is evidence that the association among the sources is significant. Focusing on the a corresponding vectors of V specifically identifies which sources agree. In cases where more than one eigenvalue is significant, the $a$ significant vectors of V provide insight into factions. When more than one set of sources is in agreement, each group of agreeing sources is considered a faction. In many cases, more than one set of sources will be in agreement with one another but not necessarily with another set of sources; each group that is in agreement would be considered a faction. Using the a significant vectors of U^* provides different but equally important results. In many cases, the elements that are being ranked can be subdivided into equivalence classes. An equivalence class is defined as subpopulations of ranked elements that are similar to one another but dissimilar from other classes. When these classes exist, U^* provides insight as to how many classes and which elements belong in each class. In summary, the R-A SVD method gives the user the ability to assess whether there is any underlying association among multiple rank sources. It then identifies when sources agree and allows for more useful and careful interpretation when analyzing rank data. / Statistics

Le ranking multipartite est un problème d'apprentissage statistique qui consiste à ordonner les observations qui appartiennent à un espace de grande dimension dans le même ordre que les labels, de sorte que les observations avec le label le plus élevé apparaissent en haut de la liste. Cette thèse vise à comprendre la nature probabiliste du problème de ranking multipartite afin d'obtenir des garanties théoriques pour les algorithmes de ranking. Dans ce cadre, la sortie d'un algorithme de ranking prend la forme d'une fonction de scoring, une fonction qui envoie l'espace des observations sur la droite réelle et l'ordre finale est construit en utilisant l'ordre induit par la droite réelle. Les contributions de ce manuscrit sont les suivantes : d'abord, nous nous concentrons sur la caractérisation des solutions optimales de ranking multipartite. Le deuxième thème de recherche est la conception d'algorithmes pour produire des fonctions de scoring. Nous proposons deux méthodes, la première utilisant une procédure d'agrégation, la deuxième un schema d'approximation. Enfin, nous revenons au problème de ranking binaire afin d'établir des vitesse minimax adaptives de convergences. / Multipartite ranking is a statistical learning problem that consists in ordering observations that belong to a high dimensional feature space in the same order as the labels, so that the observations with the highest label appear at the top of the list. This work aims to understand the probabilistic nature of the multipartite ranking problem in order to obtain theoretical guarantees for ranking algorithms. In this context, the output of a ranking algorithm takes the form of a scoring function, a function that maps the space of the observation to the real line which order is induced using the values on the real line. The contributions of this manuscript are the following : First, we focus on the characterization of optimal solutions to multipartite ranking. The second research theme is the design of algorithms to produce scoring functions. We offer two methods, the first using an aggregation procedure, the second an approximation scheme. Finally, we return to the binary ranking problem to establish adaptive minimax rate of convergence.

The Plasmodium vivax Duffy Binding Protein (DBP) is the ligand in the major pathway for P. vivax invasion of human reticulocytes, making it an appealing vaccine candidate. Region II of DBP (DBP-RII) is the minimal portion of the ligand that mediates recognition of the Duffy Antigen Receptor for Chemokines (DARC receptor) on the reticulocyte surface and constitutes the primary vaccine target. Analysis of natural variation in the coding sequences of DBP-RII revealed signature evidence for selective pressure driving variation in the residues of the putative receptor-binding site. We hypothesize that anti-DBP immunity in P. vivax infections is strain-specific and hindered by polymorphic residues altering sensitivity to immune antibody inhibition. To comprehend the human IgG response following P. vivax infections we investigated the specificity of IgG in Pursat Province, Western Cambodia. Using ELISAs, we quantified the antibody titer against five variant alleles of DBP-RII. We also sequenced the DBP-RII of the field isolates to determine their relationship to the variant alleles used in the ELISAs. When correlating the IgG titer between the DBP variants a strain-specific immune response was observed in patients with a high antibody titer to DBP-RII_AH as compared to the other variants. This was different from the correlation of high antibody titers between DBP-RII_P and DBP-RII_7.18 (ρ=0.88, p-value<0.0001) and DBP-RII_P and DBP-RII_O (ρ=0.87, p-value<0.0001). There appeared to be little correlation between specific polymorphic residues and IgG titer. Understanding the immune response to the polymorphisms within PvDBP will allow further identification of epitopes to enable the production of a more effective P. vivax vaccine

Objective of this Master's thesis is to evaluate different ways of measuring financial performance of enterprise. Economic Value Added is compared with older, traditional indicators of profitability. Emphasis lies on general applicability. In theoretical part, history of entrepreneurial goals is in addition to the actual analysis of tested indicators outlined. EVA, profitability ratios are assessed using a sample of 100 Czech manufacturing companies. In practical part, outputs of thesis apart from actual testing are stated.

The objective of this thesis is to analyse theconnection between test results and field claims of ECUs (electronic controlunits) at Scania in order to improve the acceptance criteria and evaluatesoftware testing strategies. The connection is examined through computation ofdifferent measures of dependencies such as the Pearson’s correlation, Spearman’srank correlation and Kendall’s tau. The correlations are computed from testresults in different ECU projects and considered in a predictive model based onlogistic regression. Numerical results indicate a weak connection between testresults and field claims. This is partly due to insufficient number of ECUprojects and the lack of traceability of field claims and test results. Themain conclusion confirms the present software testing strategy. Continuoussoftware release and testing results in a lower field claim and thus a betterproduct.

This dissertation is organized into three independent chapters. In Chapter 1, I consider the selection of weights for averaging a set of threshold models. Existing model averaging literature primarily focuses on averaging linear models, I consider threshold regression models. The theory I developed in that chapter demonstrates that the proposed jackknife model averaging estimator achieves asymptotic optimality when the set of candidate models are all misspecified threshold models. The simulations study demonstrates that the jackknife model averaging estimator achieves the lowest mean squared error when contrasted against other model selection and model averaging methods. In Chapter 2, I propose a model averaging framework for the synthetic control method of Abadie and Gardeazabal (2003) and Abadie et al. (2010). The proposed estimator serves a twofold purpose. First, it reduces the bias in estimating the weights each member of the donor pool receives. Secondly, it accounts for model uncertainty for the program evaluation estimation. I study two variations of the model, one where model weights are derived by solving a cross-validation quadratic program and another where each candidate model receives equal weights. Next, I show how to apply the placebo study and the conformal inference procedure for both versions of my estimator. With a simulation study, I reveal that the superior performance of the proposed procedure. In Chapter 3, which is co-authored with my advisor Professor Youngki Shin, we provide an exact computation algorithm for the maximum rank correlation estimator using the mixed integer programming (MIP) approach. We construct a new constrained optimization problem by transforming all indicator functions into binary parameters to be estimated and show that the transformation is equivalent to the original problem. Using a modern MIP solver, we apply the proposed method to an empirical example and Monte Carlo simulations. The results show that the proposed algorithm performs better than the existing alternatives. / Dissertation / Doctor of Philosophy (PhD)

As novas biotecnologias em reprodução, como a transferência de embriões e a fecundação in vitro, de certo modo, despertaram nos produtores e pesquisadores da área de melhoramento animal o interesse na seleção de fêmeas, antes pouco explorada dada a grande ênfase empregada na seleção e avaliação de reprodutores machos. A cada ano é percebido que a pecuária de corte funciona como uma empresa e, como tal, deve-se preocupar com todos os detalhes e não enfocando apenas um, como o mais importante, e é com isso que o Programa de Melhoramento Genético da Raça Nelore (PMGRN-USP) estuda cada vez mais um número maior de características como, stayability e suas possíveis aplicações à seleção das vacas Nelore, pois tem grande importância econômica. Com o objetivo de analisar a característica stayability sob enfoque Bayesiano, em registros de produção de vacas da raça Nelore, inicialmente foram implementados dois tamanhos de cadeia de Gibbs (225 ou 550 mil), dois períodos de descarte amostral (25 ou 50 mil) e duas formas de tomadas de amostra (a cada 1000 ou 250 rodadas). Os registros foram codificados como 0 (fracasso, ou menos de três partos até os 6 anos de idade) ou 1 (sucesso, ou pelo menos três partos até os 6 anos de idade) e os arquivos sofreram restrição ou não para NEP (Número Efetivo de Progênie), para número de touros dentro do grupo de contemporâneos (GC) e exclusão de grupos inteiros com média para a característica igual a 0 ou 1 (ou seja, sem variabilidade dentro do GC). Testaram-se três definições para GC. Utilizou-se o software MTGSAM for threshold (Multiple-Trait Gibbs Sampler for Animal Models), sob modelo unicaráter de touro-avô materno, para obtenção de componentes de (co)variância, estimativas de herdabilidade e soluções para cada touro (que originam as Diferenças Esperadas na Progênie - DEPs). As análises resultaram em amostras com baixa correlação serial, mostraram pequenas diferenças entre as estimativas pontuais de herdabilidade e alta correlação de rank para as DEPs dos 4180 touros avaliados. Adotou-se a implementação tamanho da cadeia/descarte amostral/amostragem como 225mil/25mil/1000 para as análises subseqüentes. A comparação entre estimativas de herdabilidade obtidas sob modelo de limiar e sob modelo linear não mostrou vantagens do primeiro, sendo que análises sob modelo linear, que têm reduzido tempo de processamento, poderiam ser preferidas quando houver muitos registros de produção, bastando transformar as estimativas de herdabilidade para escala subjacente. As comparações, sob modelo linear ajustadas para escala subjacente, entre a stayability padrão (0 ou 1) e a alternativa (considerando o número de partos até os seis anos de idade para aquelas vacas antes codificadas com valor 1), mostraram que podem ocorrer alterações na classificação de um número significativo de touros avaliados, mesmo sendo nas posições intermediárias do rank, talvez pela ligeira capacidade da característica alternativa em detectar variabilidade entre touros. Classificaram-se os 4180 touros em ordem decrescente e foram estudadas as genealogias dos 42 melhores (também chamados de TOP 1% ou aqueles que apresentam DEP para stayability superior a 57,6%) para identificar famílias importantes e avaliar a variabilidade genética da stayability. Além do touro Karvadi Imp. (essencialmente presente como avô, bisavô ou tataravô das mães dos touros TOP1%), outros genearcas com grande representatividade entre os TOP1% foram os touros Godhavari Imp. (via Kurupathy e Neofito), Rolex (via Cardeal), Rastã e Falo da BV (estes últimos pela via materna). / The new reproduction technologies like embryo transfer and in vitro fertilization brought to animal breeding researchers, and producers, a growing interest in the selection of females, which was a subject not well explored before, when the focus used to be only on evaluation and selection of males. The beef cattle industry is becoming more competitive each year, and the producers must think about all the aspects and details that can affect the production process. The PMGRN-USP (Nelore Breeding Program of the São Paulo University) studies a large number of traits to be used as selection criteria. Among these traits, there is stayability, which has a great economic importance. The aim of this project was to analyze the trait stayability, with a Bayesian approach, in a Nelore cattle population. Firstly, the implementation used two lengths of Gibbs chain (225 or 550 thousand), two periods of burn-in (25 or 50 thousand) and two thinning intervals (at each 1000 or 250 rounds). The cows were classified as 0 (failure, or less than three calves until six years of age), or 1 (success, or at least 3 calves until six years of age). The data were, or were not, restricted for NEP (Effective Number of Progeny), for number of sires in the contemporary group (GC), and for lack of variability in the contemporary group. Three different definitions of GC were tested. The software MTGSAM for threshold (Multiple-Trait Gibbs Sampler for Animal Models) was used, under an univariate sire-maternal grandsire model to get the (co)variance components, the heritability estimate and the solutions to each sire (that are used to get the Expected Progeny Differences - EPD). The results showed a low serial correlation in the samples, small differences among heritability estimates and a high rank correlation among the EPD estimates of the 4180 sires evaluated. The implementation 225000/25000/1000 was adopted to the subsequent analysis. The comparison between the heritability estimates obtained under the threshold model and the linear model didn’t show any advantage to the first. The analysis under linear models could be preferred because of its reduced processing time in large data sets, needing only a transformation of the heritability estimates to the underlying scale. The comparisons, under linear model adjusted to the underlying scale, between the standard stayalility (0 or 1) and the alternative (obtained from the perception of the fertility differences among the cows classified as 1) showed a variation in the position of a considerable number of sires in the rank, maybe because the alternative trait gets some additional variation among sires. 4180 sires were classified in a rank and the genealogies of the 42 best sires (or TOP 1%, or the ones with stayability EPD greater than 57,6%) were studied to identify major families and evaluate the genetic variation of stayability. Besides of Karvardi Imp (a bull very present as an ascendant (2, 3 or 4 generations) of the dams of the TOP1% sires), another founder sires with influence among the TOP1% were Godhavari Imp (via Kurupathy and Neofito), Rolex (via Cardeal), Rastã, and Falo da BV (the last two via maternal).

在國內外實證文獻中，已有許多學者研究市場過度反應的現象，但由採取的研究方法不盡相同，故結論也有所差異．本研究是希望能對台灣股市的個別股票報酬率作一較完整的檢視，先瞭解股票報酬率前後期的相關情形，試圖找出一些一致性，再提出股市交易的策略，並比較策略的獲利性。 將樣本分成對稱與不對稱的形成期及檢定期，採Spearman等級相關係數法先對台灣股市同一股票的報酬率在不同期間（形成期及檢定期）的表現是否有相關作一檢視，接著檢定此相關係數時間序列是否具有隨機的特質，而後將證券交易稅及手續費納入考慮後，比較三種投資策略－買入持有輸家、中間、贏家的獲利性、風險及績效表現。 本文的實證結果：由單位風險報酬率來看，投資股市應以買入有長期的策略才會有較好的績效表現，而買入持有短期的績效是最差的。以不同的樣本期間討論台灣股市是否有價格反彈，結果並不相同。在1980年-1998年的Spearman等級相關係數多為正值，表示市場在288天以下多沒有價格反彈，且以連檢定的結果多為顯著，以Jensen α檢定之，贏家投資策略能獲得超額報酬。在1990年-1998年不論是重覆取樣、未重覆取樣，Spearman等級相關係數，多為負值，表示市場應存有價格反彈的現象，且以連檢定的結果多為顯著，但以Jensen α檢定之，輸家卻無法獲得超額報酬。可能是因為考慮了交易稅與手續費，而影響了投資策略的獲利性。 / There have been many articles discussing overreaction. Because of the difference of methods and samples, the conclusions are different. This thesis tries to make a more complete examination of Taiwan Stock Market. We divide sample period to be formation period and test period which are symmetric and asymmetric. Besides that, we adopt overlapping and nonoverlapping sampling. The sample period is 1980 January 1 to 1998 January 22. Three investment strategies are buying and holding loser portfolio、middle portfolio and winner portfolio. We use Spearman rank correlation to discuss whether the return of Taiwan Stock Market has correlation between formation period and test period. Then we adopt one of nonparameter statitics analysis－run test to examine whether the time series of Spearman rank correlation is a random walk. Following are our summaries： 1.The longer period we hold the stock,the better return we acquire. 2.In the first sample (1980 Jan 1 to 1998 Jan 22), Spearman rank correlation. is almost positive, and the hypothesis of run test is significant. We imply buying and holding the winner portfolio is the best strategy. We can prove this by using Jense α. In this case, buying and holding winner can get excess return. 3.In the second sample(1990 Nov 1 to 1998 Jan 22), Spearman rank correlaion is almost negative, and the hypothesis of run test is signficant. We imply buying and holding the loser portfolio is the best strategy. But we can not prove the by using Jense α. As we can not acqure excess return by buying and holding loser portfolio.