Le domaine de la reconnaissance automatique du locuteur (RAL) recouvre l’ensembledes techniques visant à discriminer des locuteurs à partir de leurs énoncésde voix. Il se classe dans la famille des procédures d’authentification biométrique del’identité. La reconnaissance du locuteur a connu ces dernières années une avancée significativeavec un nouveau concept de représentation de l’énoncé de voix, désignésous le terme de i-vector. Ce type de représentation s’appuie sur le paradigme de modélisationpar mélange de gaussiennes et présente la particularité de se réduire numériquementà un vecteur de dimension faible, au regard des représentations précédentes,et pourtant très discriminant vis à vis du locuteur.Les travaux présentés dans cette thèse s’inscrivent dans ce nouveau contexte. Orientésautour de cette représentation, ils visent à en comprendre et évaluer les hypothèses,les points fondamentaux, le comportement et les limites.Nous avons en premier lieu conduit une analyse statistique sur cette nouvelle représentation.L’étude a porté sur l’effet et l’importance relative des différentes étapes deconstitution et d’exploitation du concept. Cette analyse a permis de mieux comprendreses caractéristiques, mais aussi de faire apparaître des défauts de la représentation quinous ont conduits à mettre en place de nouvelles transformations dans cet espace. L’objectifde ces techniques est de faire converger les données vers des modèles théoriques,à meilleur pouvoir discriminant. Nous recensons et démontrons un certain nombre depropriétés induites par ces transformations, qui justifient leur emploi. En terme de performance,ces techniques réduisent d’un ordre de grandeur de 50% les taux d’erreurdes systèmes basés sur les i-vectors et des postulats gaussiens, permettant notammentd’atteindre par la voie du cadre probabiliste gaussien les meilleurs taux de détectiondans le domaine.Une évaluation générale des composants de la méthode est ensuite détaillée dansce document. Elle met en avant l’importance de certaines étapes, permettant ainsi dedégager, par comparaison à des méthodes alternatives, les approches fondamentalesqui confèrent au concept une valeur de paradigme. Nous montrons la primauté decertaines étapes stratégiques dans la chaîne des traitements, parmi lesquelles les transformationsque nous avons mises en place, et leur relative indépendance aux méthodes et hypothèses adoptées.Des limites de la solution sont mises au jour et exposées dans une étude dite d’anisotropie,qui relativise sa capacité à produire une paramétrisation linéaire globale des variabilitésqui soit optimale.En parallèle de ces investigations, nous avons participé à l’exploration d’un nouveaumodèle alternatif à la solution la plus usuelle de représentation des énoncés devoix. Conçu par J.F. Bonastre, il produit des vecteurs sous forme de clés binaires etfournit les moyens de les comparer, en suivant une voie semi-paramétrique basée surune nouvelle approche de la problématique. Cette exploration a contribué à l’améliorationde ce modèle et à l’ouverture de nouvelles pistes. Elle a été également utile à notreévaluation du concept de i-vector.Les travaux présentés dans ce document contribuent à l’amélioration de ce modèleet à l’ouverture de nouvelles pistes. Ils sont également utiles à notre évaluation duconcept de i-vector.Enfin, quelques aménagements des solutions i-vectors à des cas particuliers ont étémis en place : nous proposons de nouvelles variantes pour gérer la décision sur lesénoncés de courte durée (qui constituent l’un des enjeux actuels du domaine) et sur lesénoncés présentant une divergence a priori (support, durée, langue distincts).L’ensemble de ces travaux vise à mieux circonscrire les pistes de recherche les plusporteuses autour de ce nouveau concept de représentation de la voix humaine / The speaker recognition field covers all the techniques intended to authentify theidentity by using voice utterances. Speaker recognition has experienced in recent yearsa significant step forward with a new concept of representation, referred to as the ivector. This type of representation is based on the Gaussian mixture model paradigmand has the distinguishing feature of being a small size vector compared to previousrepresentations, yet very discriminating towards the speaker.The works presented in this thesis are within that new context. Focused on thisrepresentation, they aim to better understand it and assess its assumptions, highlightits key points, its behaviors and limits.We first carried out a statistical analysis of this new representation. This analysishelped to better understand its characteristics, but also reveal defaults of the representationthat led us to develop new transformations. The goal of these techniques is tomove data towards a theoretical model, having a better accuracy for discrimination.We identify and demonstrate a number of properties of these transformations whichjustify their relevance. In terms of performance, applying these techniques reduce byan order of magnitude of 50% the error rate of systems based on i-vectors and Gaussianassumptions and yield the best detection rate in the field through the Gaussianprobabilistic framework. A complete evaluation of the system components is detailed later in this document.By comparing the fundamental approaches to alternative methods, this evaluationidentifies and highlights the fundamental steps that give the concept a value ofparadigm.We show the primacy of some strategic steps in the process chain, includingour propositions, and their relative independence from methods and assumptions.Limits of the solution are uncovered and exposed in a study of "anisotropy", whichreveals some lack of compliance of i-vector distributions with Gaussian assumptions.Alongside these investigations, we participated in the exploration of a new model,alternative to the most usual statistical representations of utterances, which relies on asemi- parametric representation. Designed by J.F. Bonastre, it produces binary key vectorsand provides the means to compare them. This exploration has contributed to the improvement of this model and opens new gates. It was also helpful to our evaluationof the concept of i -vector.Some adaptations of i-vector approach to special speaker recognition tasks are described: we propose new variants to handle short duration utterances ( which is oneof the current issues in the field ) and to deal with a priori mismatch (for example ofsupport, time or distinct language).We hope that this work will better highlight some of the most promising slopes ofresearch around this new concept of representation for speaker recognition

My diploma thesis focuses on the comparison of possibilities of the statistical software SAS and SPSS in the area of the factor and cluster analysis and the multidimensional scaling. They deal with the methods for identifying groups of the similar statistical values (variables). The ascertained relations among the variables can serve to decrease the proportion vectors of the variables, which describe the individual monitored objects (statistical units), which helps us to apply other various methods, for example the regression or discriminant analysis. By one of the ways for finding the similarity of variables in the cluster analysis or the multidimensional scaling is searching for their relations. Whereas the base of the factor analysis is the formulation of the relation between two variables by means of the covariances, eventually Pearson correlation coefficient, it is possible to use also coefficients of correlation for the cluster analysis and the multidimensional scaling, in some case other measures. The thesis describes mainly the command syntax of the procedures implemented in SAS and SPSS. The meaning of the individual parametres and the partial specifications of each command are explained. The results gained by various types of analyses are compared on the basis of the real dataset. The possibilities of the statistical software SAS and SPSS are evaluated in the conclusion and it is referred to their advantages or disadvantages. The attention is also paid, for example, to the form of the input dataset, to the quaility of outputs or to the partial methods.

This thesis is focused on analysis of the factors that influence the yields of the wheat. This thesis is divided into three parts. The first part opens the problem of wheat cultivation. The second one concerns the methodologies of creating the econometrics models and the third one solves the problem as a whole. Considering a large data file I have a need to reduce it by the factor analysis. I estimate relevant econometric model by application different econometrics methods. This model will show the influences of technological, soil and climatic factors on the yields of wheat. At the end I confront the observed variables with predicted ones by the graininess of soil, climate and the year of the crops.

在這科技飛騰的時代，智慧型手機使用日益普及，根據eMarketer於2016年公布台灣使用智慧型手機人口佔總人口73.4%，相較於新加坡71.8%與南韓70.4%的使用率，此比率高居全球之冠，各行業該如何運用智慧型手機市場為自己駐足的行業開創佳績，成為現今人們廣為關注的話題。 本論文研究所用之資料取自「科技部傳播調查資料庫第一期第三次（2014）：媒體的娛樂與社交功能」一般民眾（18 歲以上）之問卷資料。首先對樣本基本資料結構與特性進行描述，接著將智慧型手機使用的相關題項找出，並進行因素分析找出因素構面作為分群變數，藉由兩階段分群法進行分群，探討其各群間相關之特性與智慧型手機使用之目的。爾後從性別、年齡與教育程度等基本人口變項進行分析，進一步了解不同人口基本結構智慧型手機之使用目的之差異情形，並將「網路素養」、「社交媒體」等相關題組進行因素分析，萃取出重要共同因素後並予以命名，以探討不同媒體社交功能使用情形與智慧型手機使用目的之相關性，最後將人口基本結構與共同因素視為變數，分別採用CART、C5.0、QUEST與CHAID四種決策樹分析方法對「集群一」、「集群二」智慧型手機高度使用者進行模型之建構，使各行業可針對欲探討之集群提出行銷方針。 / With the rapid development of technology, the Internet and mobile phones play an important role in our lives. According to eMarketer 2016, 73.4% of Taiwan's population use smartphones, compared to 71.8% in Singapore and 70.4% in South Korea , Taiwan tops the list of the world. How to create success by using smartphone market is an important issue today. The data used in this thesis was taken from the Ministry of Science and Technology Survey in 2014. The survey topic was media entertainment and social functions, based on general public who are 18 years old or older. First, the structures of the sample are described. Next, we extract factors by using factor analysis. The factors are used as the cluster variables. This study uses two-stage method to cluster and explore characteristics of the relevant groups for the smartphone usage. Then, we analyze demographic variables to understand different populations of smart phones usage, and extract common factors of "Internet Literacy" and "Social Media" by using factor analysis. Finally, the basic structure of the population and the common factors are used to classify smartphone users, which helps to provide marketing guidelines.

In order of measuring the relationship between Tinder intensity and self-esteem, the Tinder Intensity Scale was developed in this present study. The scale was developed with help from exploratory factor analysis (EFA) measuring dimensionality and Cronbach's alpha testing the interrelations between the scale items. The EFA found that the Tinder Intensity Scale (Cronbach’s alpha 0.75) consists of two dimensions, one measuring frequency and emotional connection and the other measured dating behavior, such as being the first one to contact a match. The index-scores were used to measure the relationship between the self-esteem score and Tinder Intensity Scale and the scales’ two dimensions. The findings of this study suggest that the relationship between self-esteem and Tinder intensity is rather weak for both men and women, but slightly higher for men’s dating behavior and self-esteem. A reason for this could be due to the low intensity on Tinder in this sample. For future research, it is suggested to examine the relationship on a larger sample. Another suggestion is to determine causality between Tinder use and self-esteem by doing a longitudinal study, where self-esteem is measured pre-Tinder usage and after using Tinder.

Population-based evidence suggests that the prevalence of psychotic symptoms varies with ethnicity. However, the validity of these self-reported experiences, both as measures of psychotic symptoms per se and as measures of ethnic variation in risk for reporting these symptoms, is questionable. Cultural background and social position may affect the experience and expression of symptoms, which may in turn affect their measurement in survey research. This thesis explores these issues by assessing the psychometric properties of psychosis screening instruments as measures of ethnic differences in psychosis risk in two contexts, Great Britain and the United States. Following this, differences in risk for reporting psychotic symptoms were examined in the context of the varying social and economic conditions to which different ethnic groups and migrant generations are exposed. This was accomplished in the following way: Firstly, the prevalence of self-reported psychotic symptoms was examined across ethnic and generational groups in two large samples of British and American populations; Secondly, the construct validity of the instruments was assessed by means of exploratory and confirmatory factor analyses of response to their symptom indicators. The concurrent validity of the symptoms with measures of self-perceived cognitive and social dysfunction was then examined in a structural equation modeling framework; Thirdly, the cross-ethnic and cross-generational validity of response to these screening instruments was assessed in a multiple-group confirmatory factor analysis framework in which in which the psychometric characteristics of the instruments were compared; And finally, ethnic and generational differences in risk for reporting psychotic symptoms were examined in the context of differential exposure to racial discrimination. Both instruments had construct validity in their measurement of psychotic symptoms, with the exception of one item in the British instrument, as well as concurrent validity with measures of self-perceived cognitive or social dysfunction. The British instrument performed adequately in the measurement of psychotic symptoms across ethnic groups, but did not across migrant generations. The American instrument performed adequately in the measurement of psychotic symptoms across generations, but not across ethnic groups. The effect of measurement noninvariance on the estimation of risk across groups was, however, modest in size. Finally, there was evidence for risk of reporting psychotic symptoms being raised among those who were exposed to racial discrimination. These findings suggest that these self-reported psychotic symptoms constitute clinically relevant phenomena which appear phenotypically similar to the clinical symptoms in diagnosed psychotic disorder. Risk is distributed unevenly over ethnic groups in Britain and the US, is higher among minorities who were exposed to social adversity, and higher among the second generation (in most cases). These patterns are, therefore, highly suggestive of social causation in the aetiology of these self-reported symptoms.

Reports from the IPCC have been consistent in their findings: climate change is happening and human activity is the cause. The temperature of the earth’s climate has been steadily rising since the industrial revolution, with profoundly negative consequences for the natural environment. Britain is amongst the top 10 global contributors towards climate change, producing more CO2 per capita than China, and yet little is known about the relationship the British public have with the natural environment. Drawing upon DEFRA’s 2009 Survey of Public Attitudes and Behaviours Towards the Environment, a nationally representative sample of the UK, this study aims to (1) explore environmental attitudes in the DEFRA sample; (2) identify the types of environmental concern that exist in the UK and; (3) examine how environmental concern is associated with pro-environmental behaviours. The overall goal is to develop a better understanding this attitude-behaviour relationship. The thesis has 3 main findings. First, environmental concern is formed of three environmental attitudes: (a) a cognitive appraisal of plant and animal welfare (ecocentric attitude); (b) welfare of the human race (human-centric attitude); and (c) a prioritisation of the self, alongside dismissal of environmental problems (denial).Second, members of the British public can be assigned to one of four groups based on their environmental concern: Pro-environment, Neutral, Disengaged and Paradoxical (the latter 2 groups are apathetic towards environmental issues though in different ways).Third, when examining behaviour variation across these environmental concern groups, it was found, unsurprisingly, that membership of the pro- environmental group is strongly predictive of pro-environmental behaviour. What was surprising was that pro-environmental concern predicts a variety of behaviours, both easy and challenging (i.e. easy behaviour such as recycling household waste as well more challenging behaviour such as an increase use of public transportation over driving), whereas previous studies have typically found such behaviours to be unaffected by attitudes. Membership of the Neutral group also predicts pro-environmental behaviours, although this relationship is weaker and exists for fewer measures of behaviour. Disengaged and Paradoxical forms of concern are not significant predictors of behaviour. Upon examining the effect of socio-economic status (SES) on group membership and this attitude-behaviour relationship, it was found that SES does not moderate the attitude-behaviour relationship, but it does influence group membership. Respondents with higher SES were more likely to belong to neutral or pro-environment groups. After reviewing these findings, it is concluded that environmental attitudes do clearly predict behaviour, but a large portion of the UK population do not possess environmental attitudes strong enough to do so (the Disengaged and Paradoxical groups amount to 36% of the population). Future studies should focus on these apathetic groups in an attempt to understand them, determine effective methods of engagement and identify factors that increase the probability of members transitioning out of these groups.

Performance improvement intervention begins with assessment. How that assessment is interpreted can mean the difference between success and failure. Previous research of 360-degree feedback instruments has tried to reconcile the differences between multiple rater groups. Rather than searching for agreement, this research proposes to understand the meaning of the differences using multirater feedback. Individuals determine ratings based upon their own perspective and building upon the understanding of rater perspective may result in improved assessments. Data from an existing data set was processed using a second-order CFA in structural equation modeling. Covariance between the second-order factors and rater groups determined the difference in how each rater group perceived the leader.

The Rey-Osterrieth Complex Figure Task (ROCFT) has been a standard in neuropsychological assessment for six decades. Many researchers have contributed administration procedures, additional scoring systems and normative data to improve its utility. Despite the abundance of research, the original 36-point scoring system still reigns among clinicians despite documented problems with ceiling and floor effects and poor discrimination between levels of impairment. This study is an attempt to provide a new method based upon item response theory that will allow clinicians to better describe the impairment levels of their patients. Through estimation of item characteristic curves, underlying traits can be estimated while taking into account varying levels of difficulty and discrimination within the set of individual items. The ultimate goal of the current research is identification of a subset of ROCFT items that can be examined in addition to total scores to provide an extra level of information for clinicians, particularly when they are faced with a need to discriminate severely and mildly impaired patients.

The purpose of this study was to better understand the study behaviors and habits of university undergraduate students. It was designed to determine whether undergraduate students could be grouped based on their self-reported study behaviors and if any grouping system could be determined, whether group membership was related to students’ academic achievement. A total of 152 undergraduate students voluntarily participated in the current study by completing the Study Behavior Inventory instrument. All participants were enrolled in fall semester of 2010 at Florida International University. The Q factor analysis technique using principal components extraction and a varimax rotation was used in order to examine the participants in relation to each other and to detect a pattern of intercorrelations among participants based on their self-reported study behaviors. The Q factor analysis yielded a two factor structure representing two distinct student types among participants regarding their study behaviors. The first student type (i.e., Factor 1) describes proactive learners who organize both their study materials and study time well. Type 1 students are labeled “Proactive Learners with Well-Organized Study Behaviors”. The second type (i.e., Factor 2) represents students who are poorly organized as well as being very likely to procrastinate. Type 2 students are labeled “Disorganized Procrastinators”. Hierarchical linear regression was employed to examine the relationship between student type and academic achievement as measured by current grade point averages (GPAs). The results showed significant differences in GPAs between Type 1 and Type 2 students at the .05 significance level. Furthermore, student type was found to be a significant predictor of academic achievement beyond and above students’ attribute variables including sex, age, major, and enrollment status. The study has several implications for educational researchers, practitioners, and policy makers in terms of improving college students' learning behaviors and outcomes.