This master thesis concerns with emotional states and gender recognition on the basis of speech signal analysis. We used various prosodic and cepstral features for the description of the speech signal. In the text we describe non-invasive methods for glottal pulses estimation. The described features of speech were implemented in MATLAB. For their classification we used the GMM classifier, which uses the Gaussian probability distribution for modeling a feature space. Furthermore, we constructed a system for recognition of emotional states of the speaker and a system for gender recognition from speech. We tested the success of created systems with several features on speech signal segments of various lengths and compared the results. In the last part we tested the influence of speaker and gender on the success of emotional states recognition.

This document describes the three methods for the detection and classification of paralinguistic expressions such as laughing and crying from usual speech by analysis of the audio signal. The database of records was originally designed for this purpose. When analyzing everyday dialogs, music might be included, so the database was extended by four new classes as speech, music, singing with music and usual speech with background music. Feature extraction, feature reduction and classification are common steps in recognizing for all three methods. Difference of the methods is given by classification process in detail. One classification of all six classes at once is proposed in the first method called straight approach. In the second method called decision tree oriented approach we are using five intuitive sub classifiers in the tree structure and the final method uses for classification emotion coupling approach. The best features were reduced by feature evaluation using F-ratio and GMM classifiers were used for the each classification part.

Europe is in the midst of its deepest crisis since the 1930s where unsustainable debt-to-GDP levels are among the most alarming issues. It is so critical that it is unsure if the Euro can be saved. The risk of moral hazard increases within EMU when governments are taking too much risk in their public debt policies due to the anticipation that ECB or other Member States would eventually bail them out. Moreover, the SGP imposes restrictions on government deficits and debts but have previously failed to enforce them. The weakness seen in the past is that no sanctions have been put in place once the limits have been breached and the SGP is therefore incredible. Previous research on common pool and debt spillovers in a monetary union point to an upward drift of public debt as countries join the EMU. Does this argument hold true? In order to find out, 25 OECD countries between the years of 1995 and 2010 are analyzed using System GMM Arellano-Bover/Blundell-Bond one-step estimator. The primary balance, the interest payments, and GDP growth are regressed respectively in order to see through what channel EMU displays its effect. One regression will cover the entire time period and another will only cover the years from 1995 to 2007 in order to isolate the effects of the current crisis. The results, based on the years over the entire time period (including the crisis) suggest that the effect of an EMU Membership goes via the Interest payments which it is connected to positively. By using the equation of debt dynamics, the fact that net debt interest payments are higher for a country within EMU indicates, all else equal, that they have on average higher levels of debt. Nevertheless, this realization might be a crisis phenomenon and the implication of this is not clear. However more importantly, the regressions based on the years of 1995 and 2007 (prior to the crisis) did not display any significant results. These results indicate that there is no significant relationship between a country’s membership in EMU and its level of debt prior to the crisis.

Exploring massive amount of user generated data with topics posits a new way to find useful information. The topics are assumed to be “hidden” and must be “uncovered” by statistical methods such as topic modeling. However, the user generated data is typically short and messy e.g. informal chat conversations, heavy use of slang words and “noise” which could be URL’s or other forms of pseudo-text. This type of data is difficult to process for most natural language processing methods, including topic modeling. This thesis attempts to find the approach that objectively give the better topics from short and messy text in a comparative study. The compared approaches are latent Dirichlet allocation (LDA), Re-organized LDA (RO-LDA), Gaussian Mixture Model (GMM) with distributed representation of words, and a new approach based on previous work named Neural Probabilistic Topic Modeling (NPTM). It could only be concluded that NPTM have a tendency to achieve better topics on short and messy text than LDA and RO-LDA. GMM on the other hand could not produce any meaningful results at all. The results are less conclusive since NPTM suffers from long running times which prevented enough samples to be obtained for a statistical test. / Att utforska enorma mängder användargenererad data med ämnen postulerar ett nytt sätt att hitta användbar information. Ämnena antas vara “gömda” och måste “avtäckas” med statistiska metoder såsom ämnesmodellering. Dock är användargenererad data generellt sätt kort och stökig t.ex. informella chattkonversationer, mycket slangord och “brus” som kan vara URL:er eller andra former av pseudo-text. Denna typ av data är svår att bearbeta för de flesta algoritmer i naturligt språk, inklusive ämnesmodellering. Det här arbetet har försökt hitta den metod som objektivt ger dem bättre ämnena ur kort och stökig text i en jämförande studie. De metoder som jämfördes var latent Dirichlet allocation (LDA), Re-organized LDA (RO-LDA), Gaussian Mixture Model (GMM) with distributed representation of words samt en egen metod med namnet Neural Probabilistic Topic Modeling (NPTM) baserat på tidigare arbeten. Den slutsats som kan dras är att NPTM har en tendens att ge bättre ämnen på kort och stökig text jämfört med LDA och RO-LDA. GMM lyckades inte ge några meningsfulla resultat alls. Resultaten är mindre bevisande eftersom NPTM har problem med långa körtider vilket innebär att tillräckligt många stickprov inte kunde erhållas för ett statistiskt test.

No description available.

No description available.

No description available.

This dissertation consists of three essays on the empirical analysis of economic growth and technology development. In particular, I consider spillover effects in different frameworks. The first chapter outlines the three topics involved and briefly discusses the motivations, methods as well as some conclusions in each of the following chapters. The second chapter considers the spillovers in economic growth and convergence. Spillovers are prevalent in nowadays' economy. I formally model the spillover effects as the interdependence of total factor productivity (TFP), and develop a model in which spillover effects of R&D through the channel of international trade make the TFPs correlated among countries. In this sense, I apply the thoughts of international trade to the economic growth framework. Empirically, I develop a three-stage generalized method of moment(GMM) to estimate the dynamic panel spatial error autoregressive model. Simulation results show that my estimator is consistent and efficient. Through counterfactual analysis, I find that there are positive spillovers through both geographic connection and trade connection. Such a positive spillover effect, however, slows down the convergence speed. Moreover, there were little spillovers in the early 1960s. Spillover effects become stronger overtime. The third chapter is about the determinants of technology development in China. What makes my paper different from others is that I take a full consideration of the spillover effects: provincial spillovers in Science and Technology (S&T) capital as well as S&T personnel, and international spillovers through trade and FDI. The most interesting point in my paper is that I consider the indirect effects of institutions on technology development. Marketization, measured by the share of state-owned enterprises (SOEs) in the economy, affects the production of technology through different channels at different stages. I use a semiparametric varying-coefficient model to account for the effects. In this paper, I find that provincial spillovers are mainly through the externalities of S&T capital stock while international spillovers occur through trade. Marketization affects the technology development through S&T capital, S&T capital spillovers and trade. Although a certain share of SOEs is necessary for technology production, the marketization process will promote the development of technology in China in the long run. The fourth chapter looks into the provincial technology spillovers from another aspect. Instead of the S&T endowment spillovers from the nearby provinces, I consider the technology transfer from the frontier province to the targeted province as well as the absorptive capacity of the targeted province itself. Two forms of technology transfer are analyzed: the technology distance due to the structural discrepancy in the patent portfolio and the technology gap because of the difference in the patent level. Through the empirical analysis, several factors contributing to patent growth, such as S&T investment, road density, international spillovers from imports and FDI, are identified. Moreover, I find that technology transfer due to the technology distance can stimulate patent growth. However, I fail to find robust evidence of technology transfer due to the technology gap, which implies that the provincial technology convergence may not exist in China. / Ph. D.

La presente tesis se centra en la problemática existente a la hora de implementar un sistema de detección o clasificación binaria cuando es necesario combinar, integrar o fusionar diversas fuentes de información que pueden ser dependientes y heterogéneas entre sí. Las técnicas de fusión de datos tratan de combinar múltiples fuentes de información para alcanzar la exactitud y precisión en la toma de decisiones que no sería posible conseguir con el uso de una sola fuente de información de forma aislada. En un sistema de detección se pueden encontrar diferentes etapas y niveles de fusión: en la etapa de pre-detección encontramos los niveles de fusión de sensores y de características, donde se combinan los diferentes flujos de muestras proporcionados por una serie de sensores o diferentes características obtenidas del procesado estos; en la etapa de post- detección, se realiza la combinación de diferentes detectores, a través de la fusión de valoraciones continuas o de decisiones individuales aportadas por cada uno de ellos. Atendiendo al tipo de datos a combinar encontramos dos grupos: fusión soft, donde se combinan datos modelados mediante variables aleatorias continuas, caracterizadas mediante sus funciones de densidad de probabilidad (PDFs), o fusión hard, asociada a la combinación de las decisiones individuales tomadas en la etapa de fusión de detectores, donde se combinan datos binarios modelados mediante variables aleatorias discretas, caracterizadas por funciones de masa de probabilidad. Se destaca la fusión de scores como un caso particular de fusión soft asociada a la fusión de diversos detectores, en donde los datos a combinar presentan buenas propiedades discriminatorias de forma aislada y se encuentran definidos en un mismo rango normalizado [0,1]. En el presente trabajo se ha realizado una completa revisión del estado del arte en cuanto a técnicas de fusión y combinación de datos aplicadas en problemas de detección donde los datos pueden ser heterogéneos y dependientes entre sí. Se realiza una revisión en mayor profundidad de la técnica de estimación de PDFs basada en la teoría de cópulas, la cual puede ser usada en la fusión óptima de datos soft. Se destaca de forma especial tanto por su novedad e incipiente uso en el campo del procesado de señal, como por su adecuación en problemas de detección, permitiéndonos modelar de forma aislada las funciones marginales de los datos y la estructura de dependencia presente entre ellos, simplificando el problema de modelado de PDFs de datos heterogéneos y dependientes. Se ha propuesto una nueva técnica de fusión soft denominada integración-a, basada en una función de media-a, la cual, sin elevar mucho la complejidad, aporta un mayor grado de flexibilidad y de adaptación, siendo capaz de mejorar las prestaciones que se pueden obtener con respecto al resto de técnicas subóptimas utilizadas comúnmente en problemas de fusión de scores heterogéneos y dependientes entre sí. Se ha derivado un novedoso método de entrenamiento basado en el criterio de maximización parcial del área bajo la curva ROC. Se han utilizado diversas bases de datos públicas para poder testear y comprobar el correcto funcionamiento de las técnicas de fusión propuestas en problemas de autentificación multibiométrica. También se han aplicado algunas de las técnicas de fusión en la mejora de un sistema de detección de eventos acústicos. Se ha propuesto un nuevo tipo de detector basado en la teoría de cópulas denominado COCD para lidiar con el problema de la detección de señal desconocida en presencia de ruido aleatorio dependiente y no Gaussiano, centrándonos en su utilización para una aplicación de detección de eventos sonoros desconocidos. También se realiza un estudio de fusión de más de un canal de audio (utilizando más de un micrófono para captar diferentes señales) como método para incrementar las prestaciones obtenidas. / Soriano Tolosa, A. (2013). Fusión de datos estadísticamente dependientes en sistemas de detección [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/34780

La présente thèse porte sur « Les Perspectives de Ciblage de l'Inflation dans les Pays de l'UEMOA » dontl'opportunité pourrait être bien saisie à la suite de la nouvelle réforme institutionnelle de l'UMOA et de laBCEAO. Cette réforme doit s'accompagner d'une stratégie opérationnelle explicite dans la mesure où sesinnovations majeures sont très proches d'une politique de ciblage de l'inflation. Pour cela, nous évaluons lapolitique monétaire de la BCEAO axée en partie sur un régime de change fixe depuis les indépendances.Ainsi, après avoir montré la nécessité pour les Banque Centrale d’ancrer les anticipations inflationnistesdans les pays de l'UEMOA et l'apport potentiel de l'annonce d'un objectif d'inflation dans la maîtrise du tauxd'inflation (Chapitre 1), nous montrons que la politique de change fixe n'a pas été à l'origine des tauxd'inflation relativement contenus dans la l'Union, mais pire, elle procure de mauvaises résultats en termesd'une croissance économique élevée et soutenue (Chapitre 2). En cela, nous recommandons pour l’Unionun cadre de type d’une combinaison d’un régime de change intermédiaire et d’une stratégie de ciblage del’inflation. C’est ainsi qu’une cible d’inflation a été déterminée et, elle apparaît différente aussi bien d’un paysà l’autre que de l’objectif de 2% annoncé par le Comité de Politique Monétaire de la BCEAO (Chapitre 3).Nous terminons en déterminant la règle de ciblage d'inflation, et montrons que la meilleure règle optimaleque les autorités monétaires devraient adopter dans un cadre de ciblage de l'inflation concilie l’ancragenominal et la régulation conjoncturelle (Chapitre 4). / This thesis focuses on evaluating the "Prospects of Inflation Targeting in the WAEMU countries" in the wakeof the new institutional reform followed by the WAMU and the BCEAO. This reform must be accompanied byan explicit operational strategy insofar as its major innovations are very close to a policy of inflation targeting.Accordingly, we assess the impact of monetary policy of the BCEAO that focused on fixed exchange ratesince independence. In this respect, we manifest the inability of the fixed exchange rate policy to anchorinflation expectations in the WAEMU countries and the potential contribution regarding the announcement oftarget in controlling inflation (Chapter 1). In addition, we demonstrate, on the one hand, that the fixedexchange rate policy was not the cause of inflation contained in the WAEMU, and on the other, it providespoor results in terms of high and sustained economic growth (Chapter 2). Consequently, we recommend aframework of combination of an intermediate exchange rate regime and an inflation targeting strategy forWAEMU countries. Hence, an inflation target has been determined and it appears to be different from onecountry to another, against the 2% target announced by the Monetary Policy Committee of the BCEAO(Chapter 3). Finally, we define the inflation targeting rule, and show that the best optimal rule that monetaryauthorities adopt in an inflation targeting framework should ideally, reconciles nominal anchor and economicregulation (Chapter 4).