The diploma thesis is focused on using of statistical data analysis as the effective tool for evaluation of characteristics and efficiency of the company, especially economic indicators. The goal will be to analyze the data by these means, to compare them and to make the conclusions and suggestions for improvement. From the knowledge of historical data and forecasting preconditions for the future the company will gain clearer image about its development and future direction.

The master’s thesis deals with the usage of statistical analysis in the evaluation of the financial situation of the chosen company. The author focuses on the analysis of economic indocators, that are used in business practice for the assessment of the company financial situation. Summarized economic data are analyzed by financial and statistical analysis. The data comparism results to recommendations that may be impemented for company improvement. On the basis of historical data and trends prognosis the company gets specific picture about future situation and the effects on the market.

<p dir="ltr">Generating images from training samples solves the challenge of imbalanced data. It provides the necessary data to run machine learning algorithms for image classification, anomaly detection, and pattern recognition tasks. In medical settings, having imbalanced data results in higher false negatives due to a lack of positive samples. Generative Adversarial Networks (GANs) have been widely adopted for image generation. GANs allow models to train without computing intractable probability while producing high-quality images. However, evaluating GANs has been challenging for the researchers due to a need for an objective function. Most studies assess the quality of generated images and the variety of classes those images cover. Overfitting of training images, however, has received less attention from researchers. When the generated images are mere copies of the training data, GAN models will overfit and will not generalize well. This study examines the ability to detect overfitting of popular metrics: Maximum Mean Discrepancy (MMD) and Fréchet Inception Distance (FID). We investigate the metrics on two types of data: handwritten digits and chest x-ray images using Analysis of Variance (ANOVA) models.</p>

<p dir="ltr">Reinforcement learning (RL) has been a rapidly evolving field of research over the past years, enhancing developments in areas such as artificial intelligence, healthcare, and education, to name a few. Regardless of the success of RL, its inherent online learning nature presents obstacles for its real-world applications, since in many settings, online data collection with the latest learned policy can be expensive and/or dangerous (such as robotics, healthcare, and autonomous driving). This challenge has catalyzed research into offline RL, which involves reinforcement learning from previously collected static datasets, without the need for further online data collection. However, most existing offline RL methods depend on two key assumptions: unconfoundedness and positivity (also known as the full-coverage assumption), which frequently do not hold in the context of static datasets. </p><p dir="ltr">In the first part of this dissertation, we simultaneously address these two challenges by proposing a novel policy learning algorithm: PESsimistic CAusal Learning (PESCAL). We utilize the mediator variable based on Front-Door Criterion, to remove the confounding bias. Additionally, we adopt the pessimistic principle to tackle the distributional shift problem induced by the under-coverage issue. This issue refers to the mismatch of distributions between the action distributions induced by candidate policies, and the policy that generates the observational data (known as the behavior policy). Our key observation is that, by incorporating auxiliary variables that mediate the effect of actions on system dynamics, it is sufficient to learn a lower bound of the mediator distribution function, instead of the Q-function, to partially mitigate the issue of distributional shift. This insight significantly simplifies our algorithm, by circumventing the challenging task of sequential uncertainty quantification for the estimated Q-function. Moreover, we provide theoretical guarantees for the algorithms we propose, and demonstrate their efficacy through simulations, as well as real-world experiments utilizing offline datasets from a leading ride-hailing platform.</p><p dir="ltr">In the second part of this dissertation, in contrast to the first part, which approaches the distributional shift issue implicitly by penalizing the value function as a whole, we explicitly constrain the learned policy to not deviate significantly from the behavior policy, while still enabling flexible adjustment of the degree of constraints. Building upon the offline reinforcement learning algorithm, TD3+BC \cite{fujimoto2021minimalist}, we propose a model-free actor-critic algorithm with an adjustable behavior cloning (BC) term. We employ an ensemble of networks to quantify the uncertainty of the estimated value function, thus addressing the issue of overestimation. Moreover, we introduce a method that is both convenient and intuitively simple for controlling the degree of BC, through a Bernoulli random variable based on the user-specified confidence level for different offline datasets. Our proposed algorithm, named Ensemble-based Actor Critic with Adaptive Behavior Cloning (EABC), is straightforward to implement, exhibits low variance, and achieves strong performance across all D4RL benchmarks.</p>

The objective of this thesis is to analyse determinants of subjective well-being in Czechia and to compare them with other empirical evidence from Czechia and abroad. Main theoretical approaches include those emphasising "psychological" factors and those emphasising factors outside of the human personality. Data from the Public Opinion Research Centre of more than 2,000 respondents from Czechia of years 2013 and 2014 were analysed statistically. Measures of so-called global and local subjective well-being were dependent variables. Independent variables include "geographical" and demographic variables and other dummies. It was confirmed that people living in more populated buildings, with a lower space mobility, older, of a lower employment status or unemployed, lower education and left-wing oriented declare usually a lower results on the subjective well-being, too. Gender and income had variable effect on the subjective well-being. Theoretical assumptions were not confirmed considering the settlement size, mode of commuting and religion.

Submitted by William Justo Figueiro (williamjf) on 2015-06-27T12:35:42Z No. of bitstreams: 1 45.pdf: 1909890 bytes, checksum: a826b23d33eab50cce7d62b7876d703e (MD5) / Made available in DSpace on 2015-06-27T12:35:43Z (GMT). No. of bitstreams: 1 45.pdf: 1909890 bytes, checksum: a826b23d33eab50cce7d62b7876d703e (MD5) Previous issue date: 2014 / Nenhuma / O objetivo principal deste estudo foi identificar associações entre os pacotes de remuneração dos agentes das empresas brasileiras listadas no Novo Mercado com os seus indicadores de desempenho financeiro. Em segundo lugar, buscou-se verificar o cumprimento de normativo que prevê a divulgação das informações relativas à remuneração desses agentes, bem como mapear a composição de tais remunerações. A governança corporativa e a teoria da agência são usadas para desenvolver o arcabouço teórico que sustenta o estudo. A pesquisa baseia-se nos dados de 100% das empresas listadas no segmento de Novo Mercado no período de 2008 a 2012. Os dados são estudados por meio de métodos estatísticos, em particular análises descritivas e de correlação. Contrariamente às expectativas, o estudo conclui que há poucas correlações estatisticamente significativas entre as remunerações dos agentes das empresas brasileiras do Novo Mercado e seus indicadores de desempenho financeiro. Ademais, observa-se uma melhora na transparência de informações pelo cumprimento da exigência de informar a composição das remunerações dos agentes. Este estudo contribui para a literatura de gestão, sobretudo a literatura que discute a remuneração como ferramenta para mitigar os problemas de agência no Brasil. / The main objective of this study was to identify correlations between the remuneration packages of agents in Brazilian public entities listed in the “Novo Mercado” with their financial performance indicators. Secondly, it aimed to evaluate the compliance with technical standard of the disclosure of information related to the remuneration of such agents as well to map the components of this remuneration. Corporate governance and Theory of Agency were used to develop the theoretical background that supports this study. The research is based on the data from 100% of entities listed in the New Market from 2008 to 2012. Data is analyzed through a variety of statistical methodologies, in particular, descriptive analyses and correlation. Against the expectations, the study concluded that there are few correlations statistically meaningful between the remuneration of the agents of the Brazilian entities in the Novo Mercado and their financial performance indicators. In addition, an improvement in the clarity of information due to the requirements of disclosing the components of agent's remuneration is noted. This study contributes to the literature in Management, in particular, to the literature that discuss remuneration as a tool to mitigate the problems of Theory of Agency in Brazil.

Thesis (Ph.D. in Biostatistics) -- University of Colorado at Denver and Health Sciences Center, 2006. / Typescript. Includes bibliographical references (leaves 72-75). Free to UCD Anschutz Medical Campus. Online version available via ProQuest Digital Dissertations;

Thesis (Ph.D. in Analytic Health Sciences, Dept. of Preventive Medicine and Biometrics) -- University of Colorado Denver, 2007. / Typescript. Includes bibliographical references (leaves 57-59). Free to UCD Anschutz Medical Campus. Online version available via ProQuest Digital Dissertations;

Submitted by Claudinei Pracidelli (cpracide@ipen.br) on 2016-04-08T12:48:16Z No. of bitstreams: 0 / Made available in DSpace on 2016-04-08T12:48:16Z (GMT). No. of bitstreams: 0 / Dissertação (Mestrado em Tecnologia Nuclear) / IPEN/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN-CNEN/SP

Submitted by Claudinei Pracidelli (cpracide@ipen.br) on 2016-04-08T12:48:16Z No. of bitstreams: 0 / Made available in DSpace on 2016-04-08T12:48:16Z (GMT). No. of bitstreams: 0 / Para atender a legislação ambiental do Conselho Nacional do Meio Ambiente (CONAMA), a Resolução CONAMA 357/2005, é necessário desenvolver metodologias que se aplicam na medição analítica de forma correta e para um controle de qualidade é necessário a aplicação da validação de metodologia. Para atender as exigências legais, o presente trabalho desenvolveu uma metodologia para a identificação de 12 elementos metálicos, tais como: Al, Ba, Ca, Cd, Cr, Cu, Fe, K, Mg, Mn, Na e Ni em águas, avaliando 14 pontos, sendo eles GU000-01 (23°46\'49.6\"S-46°47\'22\"0W), GU000-02 (23°45\'29.5\"S-46°46\'18.7\"W), GU000-03 (23°44\'52.2\"S-46°46\'13.6\"W), GU106-04 (23°44\'44.6\"S-46°45\'25.8\"W), GU000-05 (23°44\'57.5\"S-46°45\'24.2\"W), GU107-06 (23°45\'01.2\"S-46°43\'61.5\"W), GU108-07 (23°43\'64.7\"S-46°43\'42.3\"W), GU000-08 (23°42\'96.9\"S-46°43\'61.2\"W), GU109-09 (23°43\'04.6\"S-46°43\'34.0\"W), GU105-10 (23°42\'89.9\"S-46°44\'68.7\"W), GU108-11 (23°42\'53.4\"S-46°43\'44.9\"W), GU103-12 (23°41\'88.5\"S-46°44\'67.3\"W), GU102-13 (23°41\'58.0\"S-46°43\'57.3\"W), GU000-14 (23°40\'78.2\"S-46°43\'55.0\"W), distribuídos pela Represa Guarapiranga, situada no Estado de São Paulo, aplicando a metodologia validade, realizada com base no guia do Instituto Nacional de Metrologia, Qualidade e Tecnologia (INMETRO), Orientação sobre Validação de Métodos Analíticos - DOQ-CGCE-008. Foram avaliados os parâmetros: seletividade, faixa de trabalho/linearidade, limites de detecção e quantificação, tendência/recuperação, precisão, robustez e incerteza de medição. Foi utilizado a técnica de espectrometria de Emissão Óptica com Plasma de Argônio (ICP-OES). O teste de seletividade comprovou que a matriz não interfere nas curvas analíticas elaborada; a faixa de trabalho apresentou um comportamento linear, para as amostras com e sem a matriz de interesse, com um coeficiente de correlação (r), entre 0,9965 a 1,0; os limites de detecção e quantificação do método atendem aos valores máximos permitidos pela Resolução CONAMA 357/2005; com os testes de repetitividade e de recuperação o método demonstrou ser preciso e exato, além de robusto. Posteriormente foi estimado uma incerteza de medição do método. A incerteza expandida estimada variou entre 3 e 18% da concentração encontrada. A validação da metodologia permitiu a sua aplicação para a avaliação da distribuição dos 12 elementos, nas águas da represa Guarapiranga. Foram observados valores altos para Ca, Na e K, em todos os pontos de coletas analisados, evidenciando que são os elementos que fazem parte da característica geológica da área. Os elementos Fe e Al obtiveram valores acima da legislação nos pontos da Represa (G000-01, G000-02 e G000-03). Com os testes dos parâmetros para a validação, com os cálculos estatísticos aplicados, foi possível desenvolver e aplicar uma metodologia adequada para o uso pretendido. / Dissertação (Mestrado em Tecnologia Nuclear) / IPEN/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN-CNEN/SP