No description available.

Over the past two decades, a surge of interest in the area of forecasting has produced a number of statistical models available for predicting the winners of U.S. presidential elections. While historically the domain of individuals outside the scholarly community - such as political strategists, pollsters, and journalists - presidential election forecasting has become increasingly mainstream, as a number of prominent political scientists entered the forecasting arena. With the goal of making accurate predictions well in advance of the November election, these forecasters examine several important election "fundamentals" previously shown to impact national election outcomes. In general, most models employ some measure of presidential popularity as well as a variety of indicators assessing the economic conditions prior to the election. Advancing beyond the traditional, non-scientific approaches employed by prognosticators, politicos, and pundits, today's scientific models rely on decades of voting behavior research and sophisticated statistical techniques in making accurate point estimates of the incumbent's or his party's percentage of the popular two-party vote. As the latest evolution in presidential forecasting, these models represent the most accurate and reliable method of predicting elections to date. This thesis provides an assessment of forecasting models' underlying epistemological assumptions, theoretical foundations, and methodological approaches. Additionally, this study addresses forecasting's implications for related bodies of literature, particularly its impact on studies of campaign effects. / Master of Arts

The purpose of this study was to develop easy to use price forecasting models to predict broiler price one, two and three quarters in advance. A system of five equations was developed for the two and three quarter lag models and a system of four equations was developed for the one quarter lag model. All coefficients in the equations were estimated using data published by the U.S. Department of Agriculture. The models presented are true forecasting models which predict outside the data base. The models are user oriented. Examples were presented and the data base supplied so that anyone interested in using the models can easily duplicate the results. The models were analyzed for predictive ability and were compared with the ability of the futures market to forecast broiler price. Results showed all three models predicted better than no-change extrapolation. The models predicting broiler price two and three quarters in advance predicted better than the futures market. / Master of Science

[pt] Nessa dissertação um novo método para previsão da Estrutura a Termo da Taxa de Juros Brasileira - ETTJ brasileira - conhecido como Máquina de Suporte Vetorial para Regressão é investigado, comparando-o com os métodos tradicionais, tais como modelos VAR (Vetor Auto- regressivo) e ECM (Modelos de Correção de Erros). Utiliza-se além dos retornos de títulos de renda fixa, algumas variáveis macro-econômicas, que conforme sugerido no artigo de Evans e Marshall (1998) e verificado para economia brasileira no artigo de Fukuda, Vereda e Lopes (2006) melhoram a previsão dos retornos de títulos de renda fixa no longo prazo. O experimento mostra uma melhora considerável do SVR sobre os modelos tradicionais mencionados no longo prazo, atuando ainda como ótimo indicador da direção das taxas em praticamente todos os horizontes de previsão. Para tal avaliação, foram utilizados os critérios de raiz do erro quadrado médio, erro absoluto médio, simetria direcional e simetria direcional ponderada, correta tendência para cima e correta tendência para baixo além do teste U de Theil, que faz uso da raiz do erro quadrado médio para verificar se ocorre uma melhora significativa de um modelo sobre outro. Uma vez que não existe uma maneira estruturada para escolha dos parâmetros livres do SVR, a escolha dos mesmos foi feita através de uma função do software R, que faz uma pesquisa em um domínio retangular fornecido pelo usuário. A análise dos resultados mostra que SVR é uma técnica promissora para previsão dos retornos de títulos de renda fixa, sugerindo-se ainda melhorar as escolhas dos parâmetros livres do SVR uma vez que os mesmos são meios poderosos de regularização e adaptação do ruído aos dados. / [en] In this dissertation a new method for the prediction of the Brazilian Term Structure of Interest Rates - Brazilian ETTJ - known as Support Vector Regression is investigated. This is compared with the traditional methods used in this set up, such as VAR models (Vector Autoregressive) and ECM (Error Correction Models). Besides the interest rates, some macroeconomic variables are also used, as it was suggested in a work from Evans and Marshall(1998) and verified for brazilian economy in a work from Fukuda, Vereda and Lopes (2006), the inclusion of macroeconomic variables can improve the prediction of the interest rates in long term forecasts. The experiment show some improvements in using SVR in the long term in relation to the traditional methods mentioned, acting like a realy good predictor of the direction of the interest rates along the short and long term forecasts. To make these assertions, we make use of some tests like the root mean squared error, mean absolute error, directional symmetry and weighted directional symmetry, Correct Up trend and Corret Down trend besides Theil U test, which uses the root mean squared error to verify if there is some significant improvement between two models. As there is not a structured way to choose the free parameters of SVR, a function in the R software was used in order to make a grid search over a supplied parameter ranges. The analysis of the results demonstrate that SVR is a promising technique to prediction of interest rates, suggestions are also made in order to get better the choices of the free SVR parameters once they are powerful means of regularization and adaptation to the noise in the data.

Este trabalho realiza uma análise exploratória do comportamento das venda de varejo na economia brasileira. O estudo parte da identificação de variáveis econômicas relevantes na determinação das vendas de varejo. A identificação apoia-se em proposições de outros estudos e testes de causalidade. O resultado da identificação é validado por meio da construção de um modelo de previsão de vendas no varejo. A relevância da causalidade das variáveis é verificada pela comparação do desempenho do modelo multivariado em relação ao desempenho de modelos univariados. Em seguida, o trabalho aprofunda a análise da relevância das variáveis por meio da mensuração da causalidade das taxas de juros básica da economia e ao consumidor e da mensuração do tempo médio de causalidade sobre as vendas de varejo. Por fim, as vendas de varejo são decompostas em vendas setoriais e avaliadas as mensurações de causalidade e tempo de causalidade das taxas de juros sobre os segmentos de varejo. / This work conducts an exploratory analysis on the behavior of retail sales in the brazilian economy. The analysis starts by identifying relevant economic variables in determining retail sales. The identification leans on other researches and causality tests. The identification results are validated by the contruction of a forecasting model for retail sales. The relevance of causal variables is assessed by comparing the forecasting performance of univariate models against the multivariate model. The work deepens the analysis of the causal variables by measuring the causality of the basic interest rate and the consumer interest rate over retail sales and by measuring the average time of the causality. The same analysis is extended to industry sectors in the retail sales.

In addition to air pollutants index (i.e. PSI), ambient air quality can be described by atmospheric visibility since it can be observed directly by general publics. In this study, atmospheric visibility observation, meteorological parameter monitoring, and aerosol particle sampling were conducted to investigate the influences of physicochemical properties of suspended particles and meteorological parameters on atmospheric visibility. This study further applied receptor model and multiple regression linear analysis to forecast atmospheric visibility and develop strategies for improving urban visual air quality at Taipei region. Results from regular visibility observation indicated that the average visibilities were 10.30, 8.05 and 6.00 km in the directions of Tamsui, Sonshan, and Shindian, respectively. Similar trend of visibility variation was also observed for intensive observation. Further analysis of synoptic chart and regular observation data during the period of January 2007¡VMarch 2008 showed that the lowest atmospheric visibility commonly occurred whenas the weather patterns were in sequence of eastward movement of rainy areas in southern China, southerly airstream, strong northeast monsoon, circus-sluice of high pressure outflow, and weak northeast monsoon. Results from chemical analysis of suspended particles at Taipei region indicated that major water-soluble ionic species were SO42-, NO3-, and NH4+ and followed by Cl-, while major metallic content were Ca and K. Carbonaceous analysis showed that the mass ratio of OC/EC ranged from 1.65 to 1.91 for PM2.5 and from 1.37 to 1.88 for PM2.5-10. Ammonium nitrate, organic carbon, and ammonium sulfate were the major chemical species that influenced atmospheric visibility at Taipei region. In this study, we choose the averaged atmospheric visibility in Sonshan as a dependent variable and PM10, NO2, SO2, O3, relative humidity (RH), wind direction (WD), and wind speed (WS) as independent variables to establish multiple linear regression models for forecasting the atmospheric visibility. Results of statistical analysis indicated that high correlation between forecasted and observed atmospheric visibilities was observed (R=0.7167). Furthermore, atmospheric visibility forecasting models were established for various weather patterns. The accuracies of atmospheric visibility verification (September~December, 2007) and forecasting (January~March, 2008) were 91.80% and 87.97%, respectively. This study further applied SPSS stastistic software to conduct factor analysis for atmospheric visibility. Results from factor analysis of visibility indicated that the top three factors (PM10, NO2, and SO2) accounted for 71.13% of variance. Furthermore, variable correlation analysis showed that atmospheric visibility had positive correlation with wind speed and negative correlation with other variables (PM10, NO2, SO2, O3, RH, and WD). Besides, for the significant levels of £\=0.01 or £\=0.05, all variables were proven to be significantly correlated with atmospheric visibility except O3. At Taipei region, the automobile tail emission was the major emission source causing low visibility, thus the most effective strategy for improving atmospheric visibility was to reduce the mission of automobiles and the formation of secondary aerosols containing ammonium nitrate and ammonium sulfate, which could effectively increase the atmospheric visibility at Taipei region.

Baigiamajame magistro darbe nagrinėjami įmonių bankroto analizės ir prognozavimo priimant investicinius sprendimus ypatumai. Pirmoje darbo dalyje yra analizuojami investicijų ir investavimo ypatumai, išskiriant investicijų formas ir pateikiant jų klasifikaciją bei įtaką šalies įmonėms. Antroje darbo dalyje atskleidžiama bankroto esmė, analizuojamos jo atsiradimo priežastys ir požymiai, pateikiama Lietuvos įmonių bankrotų ir tyčinių bankrotų apžvalga bei analizuojama bankroto prognozavimo svarba ir prevenciniai prognozavimo metodai. Trečiojoje dalyje atliekamas praktinis šalies statybų sektoriaus įmonių vertinimas UAB „Telvalda“ ir UAB „Mažeiva“ pavyzdžiu. Šioje dalyje išskirti ir įvertinti įmonių išorės – vidaus aplinkos veiksniai, įvertinta veiklos struktūra ir turto bei nuosavo kapitalo įtaka pelnui (nuostoliui), pateikta santykinių finansinių rodiklių analizė ir įvertinta statybos sektoriaus įmonių bankroto tikimybė. Darbą sudaro 5 dalys: įvadas, teorinė dalis, UAB „Telvalda“ ir UAB „Mažeiva“ bankroto tikimybės analizė, išvados ir siūlymai, literatūros sąrašas. Darbo apimtis – 86 p. teksto be priedų, 35 pav., 19 lent., 75 bibliografiniai šaltiniai. Atskirai pridedami darbo priedai. / This final Master thesis analyses the corporate bankruptcy and forecasting in making investment decisions. The first part of the work presents the analysis of investments and investment characteristics, highlighting the investment forms and presenting their classification and their influence on the country's businesses. The second part of the work reveals the essence of bankruptcy, and analyzes its causes and symptoms, presents the overview of the Lithuanian corporate bankruptcies and fraudulent bankruptcy and analysis of the importance of bankruptcy forecasting, and preventive methods of forecasting. The third part contains the practical assessment of the national companies in the construction sector based on the cases of UAB Telvalda and UAB Mažeiva. This section identifies and assesses the company's external – internal environmental factors, evaluates the operational structure and performance, impact of assets and equity on profit (loss), presents the analysis of relative financial indicators and evaluates the bankruptcy likelihood of companies in the construction industry. The work consists of 5 parts: introduction, theoretical part, bankruptcy probability analysis of UAB Telvalda and UAB Mažeiva, conclusions and recommendations, references. Volume of the work – 86 pages of text without annexes, 35 figures, 19 tables, 75 bibliographical sources. Supplements to the work are attached separately.

Diss. (sammanfattning) Alnarp : Sveriges lantbruksuniversitet, 2006. / Härtill 4 uppsatser.

Submitted by Luciana Paro Scarin Freitas null (melscarin@gmail.com) on 2016-09-15T19:43:12Z No. of bitstreams: 1 Tese Final - Luciana Paro Scarin Freitas - 150916.pdf: 2268932 bytes, checksum: 6258cf968244fdbb360b56af8ef82a25 (MD5) / Approved for entry into archive by Juliano Benedito Ferreira (julianoferreira@reitoria.unesp.br) on 2016-09-15T19:48:07Z (GMT) No. of bitstreams: 1 freitas_lps_dr_ilha.pdf: 2268932 bytes, checksum: 6258cf968244fdbb360b56af8ef82a25 (MD5) / Made available in DSpace on 2016-09-15T19:48:07Z (GMT). No. of bitstreams: 1 freitas_lps_dr_ilha.pdf: 2268932 bytes, checksum: 6258cf968244fdbb360b56af8ef82a25 (MD5) Previous issue date: 2016-09-05 / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) / O dióxido de carbono (CO2) é considerado um dos principais gases do efeito estufa adicional e contribui significativamente para as mudanças climáticas globais. Áreas agrícolas oferecem uma oportunidade para mitigar esse efeito, uma vez que, dependendo de seu uso e manejo, são capazes de armazenar grandes quantidades de carbono, retirando-as da atmosfera. A produção de CO2 no solo é resultado de processos biológicos, como a decomposição da matéria orgânica e respiração de raízes e organismos do solo, fenômeno chamado de emissão de CO2 do solo (FCO2). O objetivo deste trabalho foi utilizar as redes neurais artificiais para estudo e previsão de padrões espaço-temporais da emissão de CO2 do solo em áreas de cana-de-açúcar em sistema de cana crua, colheita mecanizada, quando grandes quantidades de palhas são depositadas sobre a superfície do solo. Valores de FCO2 foram coletados em áreas de cultivo comercial no Sudeste do Estado de São Paulo, registrados por meio do sistema LI-8100, em gradeados amostrais para determinação da variabilidade espaçotemporal de FCO2, e atributos físicos e químicos do solo. Foram utilizados dados referentes a estudos realizados nos anos de 2008, 2010 e 2012, no período após a operação de colheita mecânica da cultura. Uma rede neural Perceptron Multi-Camadas via algoritmo backpropagation foi aplicada para estimar a emissão de FCO2 do ano de 2012, utilizando os dados referentes aos anos de 2008 e 2010 para treinamento da rede neural. A rede neural inicialmente apresentou um MAPE de 18,3852 coeficiente de determinação R2 de 0,9188. Os dados obtidos do FCO2 observado e do FCO2 estimado apresentam moderada dependência espacial, e pelos mapas do padrão espacial do fluxo de CO2 é observado que a rede neural apresentou considerável similaridade com os dados observados, identificando os pontos característicos de maior emissão como também os de menor emissão de CO2. Portanto, os resultados indicam que a rede neural artificial pode fornecer estimativas com confiabilidade para a avaliação de FCO2 a partir de dados de atributos físicos e químicos do solo, sendo capaz de caracterizar a variabilidade espaçotemporal desse atributo em áreas de cana-de-açúcar, sob o sistema de cana crua no Sudeste do Estado de São Paulo. / Carbon dioxide (CO2) is considered one of the main gases additional greenhouse effect and contributes significantly to global climate change. Agriculture areas offer an opportunity to mitigate this effect, since, depending on its use and handling, are capable of storing large amounts of carbon, removing them from the atmosphere. The CO2 production in soil is the result of biological processes such as the decomposition of organic matter and breathing roots and soil organisms, a phenomenon called soil CO2 emissions (FCO2). The aim of this study was to use artificial neural networks to study and forecast patterns spatiotemporal of soil CO2 emission in areas of sugarcane in raw cane system, mechanical harvesting, when large amounts of straw are deposited on soil surface. FCO2 values were collected in areas of commercial cultivation in southeastern of the state of São Paulo, registered through the LI-8100 system, sample grilles for determining the spatiotemporal variability of FCO2, and physical and chemical soil properties. The used data were from studies conducted in the years 2008, 2010 and 2012, in the period after the mechanical harvesting operation culture. A Multilayer Perceptron neural network with backpropagation algorithm was applied to estimate the emission of FCO2 in the year 2012, using data from the years 2008 and 2010 to the neural network training. The neural network initially presented a MAPE of 18.3852 and determination coefficient R2 of 0.9188. Data obtained from the observed FCO2 and FCO2 estimated present moderate spatial dependence, and observing the maps of the spatial pattern of the CO2 flow show that neural network presents considerable similarity to the observed data, identifying the higher and lower characteristic points of CO2 emissions. Therefore, the results indicate that the artificial neural network can provide reliability for the evaluation of FCO2 from data of physical and chemical soil properties, being able to describe the spatiotemporal variability of this attribute in sugarcane fields, under the crude cane system in the southeastern of the state of São Paulo. / CNPq: 152199/2012-8

Orientador: Anna Diva Plasencia Lotufo / Resumo: O dióxido de carbono (CO2) é considerado um dos principais gases do efeito estufa adicional e contribui significativamente para as mudanças climáticas globais. Áreas agrícolas oferecem uma oportunidade para mitigar esse efeito, uma vez que, dependendo de seu uso e manejo, são capazes de armazenar grandes quantidades de carbono, retirando-as da atmosfera. A produção de CO2 no solo é resultado de processos biológicos, como a decomposição da matéria orgânica e respiração de raízes e organismos do solo, fenômeno chamado de emissão de CO2 do solo (FCO2). O objetivo deste trabalho foi utilizar as redes neurais artificiais para estudo e previsão de padrões espaço-temporais da emissão de CO2 do solo em áreas de cana-de-açúcar em sistema de cana crua, colheita mecanizada, quando grandes quantidades de palhas são depositadas sobre a superfície do solo. Valores de FCO2 foram coletados em áreas de cultivo comercial no Sudeste do Estado de São Paulo, registrados por meio do sistema LI-8100, em gradeados amostrais para determinação da variabilidade espaçotemporal de FCO2, e atributos físicos e químicos do solo. Foram utilizados dados referentes a estudos realizados nos anos de 2008, 2010 e 2012, no período após a operação de colheita mecânica da cultura. Uma rede neural Perceptron Multi-Camadas via algoritmo backpropagation foi aplicada para estimar a emissão de FCO2 do ano de 2012, utilizando os dados referentes aos anos de 2008 e 2010 para treinamento da rede neural. A rede neural inici... (Resumo completo, clicar acesso eletrônico abaixo) / Doutor