Generalization transitions in Hidden-Layer neural networks for third-order feature discrimination

Stochastic learning processes for a specific feature detector are studied. This technique is applied to nonsmooth multilayer neural networks requested to perform a discrimination task of order 3 based on the ssT-block¿ssC-block problem. Our system proves to be capable of achieving perfect generalization, after presenting finite numbers of examples, by undergoing a phase transition. The corresponding annealed theory, which involves the Ising model under external field, shows good agreement with Monte Carlo simulations.

EU-15 sovereign governments’ cost of borrowing after seven years of monetary union

Yield spreads over 10-year German government securities of the EU-15 countries converged dramatically in the seven years after the beginning of Monetary Integration. In this paper, we investigate the relative influence of systemic and idiosyncratic risk factors on their behaviour. Our conclusions suggest that in EMU-countries the relative importance of domestic risk factors (both credit and liquidity risk factors) is higher than that of international factors, which appear to play a secondary but significant role in non-EMU countries.

Strategic practices : : hidden histories of gender in Finland 1880-2005 : projektikuvaus

Ajankohtaista

“Weathering a Hidden Storm”: An Application of Andersen’s Behavioral Model of Health, and Health Services Use for Those with Diagnosable Anxiety Disorder

Abstract: Research has primarily focused on depression and mood disorders, but little research has been devoted to an examination of mental health services use amongst those with diagnosable anxiety disorder (Wittchen et al., 2002; Bergeron et al., 2005). This study examined the possible predicting factors for mental health services utilization amongst those with identifiable anxiety disorder in the Canadian population. The methods used for this study was the application of Andersen’s Behavioral Model of Health Services Use, where predisposing, need and enabling characteristics were regressed on the dependent variable of mental health services use. This study used the Canadian Community Health Survey (cycle 1.2: Mental Health and Well-Being) in a secondary data analysis. Several multiple logistics models predicted the likelihood to seek and use mental health services. Predisposing characteristics of gender and age, Enabling characteristics of education and geographical location, and those with co-occurring mood disorders were at the greatest increased likelihood to seek and use mental health services.

"Weathering a hidden storm" : an application of Andersen's behavioral model of health, and health services use for those with diagnosable anxiety disorder

"Weathering a Hidden Storm": An App~ication of Andersen's Behaviora~ Mode~ of Hea~th, and Hea~th Services Use for Those With Diagnosab~e Anxiety Disorder Research has primarily focused on depression and mood disorders, but little research has been devoted to an examination of mental health services use amongst those with diagnosable anxiety disorder (Wittchen et al., 2002; Bergeron et al., 2005). This study examined the possible predicting factors for mental health services utilization amongst those with identifiable anxiety disorder in the Canadian population. The methods used for this study was the application of Andersen's Behavioral Model of Health Services Use, where predisposing, need and enabling 111 characteristics were regressed on the dependent variable of mental health services use. This study used the Canadian Community Health Survey (cycle 1.2: Mental Health and Well- Being) in a secondary data analysis. Several multiple logistics models predicted the likelihood to seek and use mental health services. Predisposing characteristics of gender and age, Enabling characteristics of education and geographical location, and those with co-occurring mood disorders were at the greatest increased likelihood to seek and use mental health services.

Hidden Motives: An Analysis of Online ESL Teacher Hiring Practices in Japan and Hong Kong

Hidden Motives: An Analysis of Online English as a Second Language (ESL) Teacher Hiring Practices in Japan and Hong Kong is a qualitative research paper examines and compares two large-scale Asian English language teaching programs: Japan’s Japan Exchange and Teaching (JET) Programme (JET Programme, 2010) and Hong Kong’s Native-speaking English Teacher (NET) Scheme (NET Scheme, 2013). Both government sponsored programs recruit internationally and invite participants to work within each country’s public schools while living amongst local communities and both programs utilize their online presence to attract, inform, and recruit individuals. The purpose of this research is to investigate whether the JET and NET websites are transparent with their governmental motives aside from improving their students’ English language abilities. While JET and NET websites were interrogated, the research questions were regularly revisited to determine if the two sites made any underlying motives clear to the candidates. The research, supported by academic literature, exposed the JET Programme website to be a branch of the Japanese government’s soft power campaign, whereby JET teachers were hired firstly as potential advocates for Japan and Japanese culture rather than English teachers. Conversely, the NET Scheme appeared to be solely commissioned for English language improvement as reflected by their website. Findings from the research can provide insight to applicants to help them decide if they want to participant in these programs. Without clearly understanding the background that motivates these programs, participants may unknowingly be used to support the host government’s agendas.

The Hidden Truths of Self-Presentation, Self-Disclosure, and Deception on Online Dating Profiles: A Call for Change

The way in which individuals seek romantic partners has changed considerably in the past decades, most notably through online dating sites (ODSs). Despite the possibility of misrepresentation amongst client users, such sites continue to grow in popularity because ODSs provide a large pool from which individuals can select and attract potential partners. While much research has been undertaken on ODSs, little empirical research has examined postsecondary students’ use of ODSs. Therefore, this study sought to investigate why postsecondary students have become involved with and how they present themselves on ODSs. The researcher surveyed 20 postsecondary students and conducted in-depth interviews with 2 participants who use ODSs. Although the limited sample prevented results from being generalized, quantitative and qualitative analyses suggest that participants became involved with ODSs for various purposes, such as seeking long-term relationships and/or marriage partners, or simply exploring or visiting ODSs out of curiosity. Findings indicate that ODS users’ physical appearance and/or “attractiveness” is considered the strongest predictor of relationship success. The study discusses how participants’ self-presentation affects outcomes of ODS usage, particularly when negative self-identification and presentation corresponding to factors such as individuals’ weight and age are taken into account.

Essays in open economy macroeconomics with borrowing frictions

Cette thèse comporte trois essais en macroéconomie en économie ouverte et commerce international. Je considère tour à tour les questions suivantes: sous quelles conditions est-il optimal pour un pays de former une union économique? (essai 1); l'augmentation de la dispersion transversale des avoirs extérieurs nets des pays est-elle compatible avec une dispersion relativement stable des taux d'investissement? (essai 2); le risque de perte de marché à l'exportation du fait de l'existence des zones de commerce préférentiel joue t-il un rôle dans la décision des pays exclus de négocier des accords commerciaux à leur tour? (essai 3). Le premier essai examine les conditions d'optimalité d'une union économique. Il s'intéresse à une motivation particulière: le partage du risque lié aux fluctuations du revenu. Dans la situation initiale, les pays ont très peu d'opportunités pour partager le risque à cause des frictions: les marchés financiers internationaux sont incomplets et il n'y pas de mécanisme pour faire respecter les contrats de crédit entre pays. Dans ce contexte, une union économique apparait comme un arrangement qui pallie à ces frictions entre les pays membres seulement. Cependant, l'union dans son ensemble continue de faire face à ces frictions lorsqu'elle échange avec le reste du monde. L'arbitrage clé dans le modèle est le suivant. D'un coté, l'intégration économique permet un meilleur partage du risque entre pays membres et la possibilité pour le partenaire pauvre d'utiliser la ligne de crédit du partenaire riche en cas de besoin. De l'autre coté, l'union peut faire face à une limite de crédit plus restrictive parce que résilier la dette extérieure est moins coûteux pour les membres l'union. De plus, le fait que le partenaire pauvre peut utiliser la limite de crédit du partenaire riche génère une externalité négative pour ce dernier qui se retrouve plus fréquemment contraint au niveau des marchés internationaux des capitaux. En conformité avec les faits observés sur l'intégration économique, le modèle prédit que les unions économiques sont relativement peu fréquentes, sont plus susceptibles d'être créées parmi des pays homogènes, et généralement riches. Le deuxième essai porte sur la dispersion des avoirs extérieurs nets et la relation avec la dispersion des taux d'investissement. Au cours des récentes décennies, la dispersion croissante des déséquilibres extérieurs et les niveaux record atteints par certaines grandes économies ont reçu une attention considérable. On pourrait attribuer ce phénomène à une réduction des barrières aux mouvements internationaux des capitaux. Mais dans ce cas, il est légitime de s'attendre à une augmentation de la dispersion au niveau des taux d'investissement; ceci, parce que le financement des besoins en investissements constitue une raison fondamentale pour laquelle les pays échangent les capitaux. Les données indiquent cependant que la dispersion des taux d'investissement est restée relativement stable au cours des récentes décennies. Pour réconcilier ces faits, je construis un modèle d'équilibre général dynamique et stochastique où les pays sont hétérogènes en raison des chocs idiosyncratiques à leurs niveaux de productivité totale des facteurs. Au niveau des marchés internationaux des capitaux, le menu des actifs disponibles est restreint à une obligation sans risque et il n'y a pas de mécanisme pour faire respecter les contrats de crédit entre pays. A tout moment, un pays peut choisir de résilier sa dette extérieure sous peine d'exclusion financière et d'un coût direct. Ce coût direct reflète les canaux autres que l'exclusion financière à travers lesquels les pays en défaut sont pénalisés. Lorsque le modèle est calibré pour reproduire l'évolution de la dispersion transversale des avoirs extérieurs nets, il produit une dispersion relativement stable des taux d'investissement. La raison principale est que les incitations que les pays ont à investir sont liées à la productivité. Avec l'intégration financière, même si les opportunités d'emprunt se sont multipliées, les incitations à investir n'ont pas beaucoup changé. Ce qui permet de générer une dispersion accrue de la position des avoirs extérieurs nets des pays avec une dispersion relativement stable des taux d'investissement. Le troisième essai analyse un aspect de l'interdépendance dans la formation des accords commerciaux préférentiels: j'examine empiriquement si le risque de diversion des exportations en faveur des pays membres des zones de commerce préférentiel est un facteur déterminant dans la décision des pays exclus de ces accords de négocier un accord à leur tour. Je construis un indicateur qui mesure le potentiel de diversion des exportations auquel font face les pays et estime un modèle probit de formation des zones de commerce préférentiel créées entre 1961 et 2005. Les résultats confirment que les pays confrontés à un plus grand potentiel de détournement des échanges sont plus susceptibles de former une zone de commerce préférentiel à leur tour.

Probabilistic Independence Networks for Hidden Markov Probability Models

Graphical techniques for modeling the dependencies of randomvariables have been explored in a variety of different areas includingstatistics, statistical physics, artificial intelligence, speech recognition, image processing, and genetics.Formalisms for manipulating these models have been developedrelatively independently in these research communities. In this paper weexplore hidden Markov models (HMMs) and related structures within the general framework of probabilistic independencenetworks (PINs). The paper contains a self-contained review of the basic principles of PINs.It is shown that the well-known forward-backward (F-B) and Viterbialgorithms for HMMs are special cases of more general inference algorithms forarbitrary PINs. Furthermore, the existence of inference and estimationalgorithms for more general graphical models provides a set of analysistools for HMM practitioners who wish to explore a richer class of HMMstructures.Examples of relatively complex models to handle sensorfusion and coarticulationin speech recognitionare introduced and treated within the graphical model framework toillustrate the advantages of the general approach.

Factorial Hidden Markov Models

We present a framework for learning in hidden Markov models with distributed state representations. Within this framework, we derive a learning algorithm based on the Expectation--Maximization (EM) procedure for maximum likelihood estimation. Analogous to the standard Baum-Welch update rules, the M-step of our algorithm is exact and can be solved analytically. However, due to the combinatorial nature of the hidden state representation, the exact E-step is intractable. A simple and tractable mean field approximation is derived. Empirical results on a set of problems suggest that both the mean field approximation and Gibbs sampling are viable alternatives to the computationally expensive exact algorithm.

Estimating Dependency Structure as a Hidden Variable

This paper introduces a probability model, the mixture of trees that can account for sparse, dynamically changing dependence relationships. We present a family of efficient algorithms that use EMand the Minimum Spanning Tree algorithm to find the ML and MAP mixtureof trees for a variety of priors, including the Dirichlet and the MDL priors.