No description available.

Hydrologic and climate models predict variables through a simplification of the underlying complex natural processes. Model development involves minimising predictive uncertainty. Predictive uncertainty arises from three broad sources which are measurement error in observed responses, uncertainty of input variables and model structural error. This thesis introduces ways to improve predictive accuracy of hydroclimatic models by considering input and structural uncertainties. The specific methods developed to reduce the uncertainty because of erroneous inputs and model structural errors are outlined below. The uncertainty in hydrological model inputs, if ignored, introduces systematic biases in the parameters estimated. This thesis presents a method, known as simulation extrapolation (SIMEX), to ascertain the extent of parameter bias. SIMEX starts by generating a series of alternate inputs by artificially adding white noise in increasing multiples of the known input error variance. The resulting alternate parameter sets allow formulation of an empirical relationship between their values and the level of noise present. SIMEX is based on the theory that the trend in alternate parameters can be extrapolated back to the notional error free zone. The case study relates to erroneous sea surface temperature anomaly (SSTA) records used as input variables of a linear model to predict the Southern Oscillation Index (SOI). SIMEX achieves a reduction in residual errors from the SOI prediction. Besides, a hydrologic application of SIMEX is demonstrated by a synthetic simulation within a three-parameter conceptual rainfall runoff model. This thesis next advocates reductions of structural uncertainty of any single model by combining multiple alternative model responses. Current approaches for combining hydroclimatic forecasts are generally limited to using combination weights that remain static over time. This research develops a methodology for combining forecasts from multiple models in a dynamic setting as an improvement of over static weight combination. The model responses are mixed on a pair wise basis using mixing weights that vary in time reflecting the persistence of individual model skills. The concept is referred here as the pair wise dynamic weight combination. Two approaches for forecasting the dynamic weights are developed. The first of the two approaches uses a mixture of two basis distributions which are three category ordered logistic regression model and a generalised linear autoregressive model. The second approach uses a modified nearest neighbour approach to forecast the future weights. These alternatives are used to first combine a univariate response forecast, the NINO3.4 SSTA index. This is followed by a similar combination, but for the entire global gridded SSTA forecast field. Results from these applications show significant improvements being achieved due to the dynamic model combination approach. The last application demonstrating the dynamic combination logic, uses the dynamically combined multivariate SSTA forecast field as the basis of developing multi-site flow forecasts in the Namoi River catchment in eastern Australia. To further reduce structural uncertainty in the flow forecasts, three forecast models are formulated and the dynamic combination approach applied again. The study demonstrates that improved SSTA forecast (due to dynamic combination) in turn improves all three flow forecasts, while the dynamic combination of the three flow forecasts results in further improvements.

Diss. (sammanfattning) Stockholm : Karolinska institutet, 2007. / Härtill 5 uppsatser.

Diss. (sammanfattning) Stockholm : Karolinska institutet, 2006. / Härtill 4 uppsatser.

Dissertation (Ph.D.) -- University of Texas Southwestern Medical Center at Dallas, 2006. / Vita. Bibliography: p.97-109

During multisensory integration, information from distinct sensory systems that refers to the same physical event is combined. For example, the sound and image that an individual generates as s/he interacts with the world, will provide the nervous system with multiple cues which can be integrated to estimate the individual’s position in the environment. However, the information that is perceived through different sensory pathways/systems can be qualitatively different. The information can be redundant and describe the same property of an event in a common reference frame (i.e., the image and sound referring to the individual’s location), or it can be complementary. Combining complementary information can be advantageous in that it extends the range and richness of the information available to the nervous system, but can also be superfluous and unnecessary to the task at hand – i.e. olfactory cues about the individuals perfume can increase the richness of the representation but not necessarily aid in localisation. Over the last century or so, a large body of research has focused on different aspects of multisensory interactions at both the behavioural and neural levels. It is currently unclear whether the mechanisms underlying multisensory interactions for both type of cue are similar or not. Moreover, the evidence for differences in behavioural outcome, dependent on the nature of the cue, is growing. Such cue property effects possibly reflect a processing heuristic for more efficient parsing of the vast amount of sensory information available to the nervous system at any one time. The present thesis assesses the effects of cue properties (i.e., redundant or complementary) on multisensory processing and reports a series of experiments demonstrating that the nature of the cue, defined by the task of the observer, influences whether the cues compete for representation as a result of interacting, or whether instead multisensory information produces an optimal increase in reliability of the event estimate. Moreover, a bridging series of experiments demonstrate the key role of redundancy in inferring that two signals have a common physical cause and should be integrated, despite conflict in the cues. The experiments provide insights into the different strategies adopted by the nervous system and some tentative evidence for possible, distinct underlying mechanisms.

A combinação de algoritmos tem despertado interesse para melhorar o desempenho de filtros adaptativos. Esse método consiste em combinar linearmente as saídas de dois filtros operando em paralelo com passos de adaptação diferentes para se obter um filtro com conver- gência rápida e um erro quadrático médio em excesso (EMSE - excess mean squared error) reduzido. Nesse contexto, foi proposta a combinação afim de dois algoritmos LMS (least-mean square), cujo parâmetro de mistura não fica restrito ao intervalo [0, 1] e por isso é considerada como uma generalização da combinação convexa. Neste trabalho, a combinação afim de dois algoritmos LMS é estendida para os algoritmos supervisionados NLMS (normalized LMS) e RLS (recursive least squares) e também para equalização autodidata, usando o CMA (constant modulus algorithm). Foi feita uma análise em regime da combinação afim desses algoritmos de forma unificada, considerando entrada branca ou colorida e ambientes estacionários ou não- estacionários. Através dessa análise, verificou-se que a combinação afim de dois algoritmos da mesma família pode apresentar uma redução de EMSE de até 3 dB em relação ao EMSE de seus filtros componentes e conseqüentemente ao EMSE da combinação convexa. Para garantir que a estimativa combinada seja pelo menos tão boa quanto a do melhor filtro componente, foram propostos e analisados três novos algoritmos para adaptação do parâmetro de mistura. Utilizando resultados da análise desses algoritmos em conjunto com os resultados da análise de transitório de filtros adaptativos, analisou-se o comportamento transitório da combinação afim. Através de simulações, observou-se uma boa concordância entre os resultados analíticos e os de simulação. No caso de equalização autodidata, também foi proposta uma combinação de dois equalizadores CMA com inicializações diferentes. Verificou-se através de simulações que em alguns casos a combinação afim é capaz de evitar a convergência para mínimos locais da função custo do módulo constante. / In order to improve the performance of adaptive filters, the combination of algorithms is receiving much attention in the literature. This method combines linearly the outputs of two filters operating in parallel with different step-sizes to obtain an adaptive filter with fast convergence and reduced excess mean squared error (EMSE). In this context, it was proposed an affine combination of two least-mean square (LMS) filters, whose mixing parameter is not restricted to the interval [0, 1]. Hence, the affine combination is a generalization of the convex combination. In this work, the affine combination of two LMS algorithms is extended to the supervised algorithms NLMS (normalized LMS) and RLS (recursive least squares), and also to blind equalization, using the constant modulus algorithm (CMA). A steady-state analysis of the affine combination of the considered algorithms is presented in a unified manner, assuming white or colored inputs, and stationary or nonstationary environments. Through the analysis, it was observed that the affine combination of two algorithms of the same family can provide a 3 dB EMSE gain in relation to its best component filter and consequently in relation to the convex combination. To ensure that the combined estimate is at least as good as the best of the component filters, three new algorithms to adapt the mixing parameter were proposed and analyzed. Using the analysis results of these algorithms in conjunction with the results of the transient analysis of adaptive filters, the transient behavior of the affine combination was analyzed. Through simulations, a good agreement between analytical and experimental results was always observed. In the blind equalization case, a combination of two CMA equalizers with different initializations was also proposed. The simulation results suggest that the affine combination can avoid local minima of the constant modulus cost function.

To aid in understanding headache etiology and symptomatology, psychological and physiological variables were examined in patients with migraine and combination headaches (combined migraine and muscle-contraction headaches). One hundred patients being evaluated for treatment of their headaches at The New England Center for Headache participated in this study. They were assigned to the migraine or combination group, based on diagnoses made by three headache specialists—a psychologist, a psychiatrist, and a nuerologist. Personality data from the MMPI and frontalis electromyographic readings reflecting muscle tensions across three stimulus conditions were compared between the two groups. Subjects were also asked to rate the perceived level of stress elicited by the three conditions.

A obtenção de previsões com maior acuracidade é uma necessidade constantemente requerida, em tempos onde há imensa disponibilidade de dados e recursos computacionais cada dia mais eficientes. Tais critérios possibilitaram o desenvolvimento de muitas técnicas de previsão individual ou de métodos de combinação que são considerados eficientes no intuito de reduzir erros. O desenvolvimento de novas técnicas, por sua vez, promove questionamentos quanto à identificação de quantas ou quais técnicas de previsão individual combinar. A literatura não é unânime ao tentar responder a estes questionamentos e indica a importância da correlação entre os erros de previsão na precisão da combinação. Posto isso, esta tese apresenta uma alternativa aos métodos atuais de combinar previsões, contemplando a correlação entre os erros de previsão, além de propor uma forma de identificar técnicas de previsão que sejam distintas quanto à modelagem de características da série de dados. Para identificar grupos de técnicas de previsão individual que sejam similares, utilizou-se a Análise de Agrupamentos em erros gerados por 15 técnicas de previsão que modelaram uma mesma série de dados real com tendência e sazonalidade. O resultado indicou a formação de 3 agrupamentos. Como alternativa aos métodos atuais de combinar previsão e selecionar a quantidade adequada de técnicas, utilizou-se a Análise de Componentes Principais. O método proposto mostrou-se viável quando comparado com outros métodos de combinação e quando submetido à modelagem de séries com maior variabilidade. / The obtaining of more accurate forecasts is a necessity often required in times where there is a huge availability of data and computing resources becoming more efficient every day. These criteria allowed the development of many individual forecasting techniques or combination methods that are considered efficient in order to reduce errors. The development of new techniques, in turn, promotes questioning as the identification of how many or which techniques to combine individual forecasts. The literature is not unanimous when trying to answer these questions and indicates the importance of the correlation between forecast errors on the accuracy of the combination. That said, this presents an alternative to current methods of combining forecasts, considering the correlation between forecast errors, and propose a way to identify predictive techniques that are different about the modeling features of the data series. To identify groups of individual forecasting techniques that are similar, it was used the cluster analysis on errors generated by 15 forecasting techniques that shaped the same series of real data with trend and seasonality. The result indicated the formation of 3 clusters. As an alternative to current methods of combining forecasting and selecting the appropriate amount of techniques, it was used the Principal Component Analysis. The proposed method has proved feasible when compared to other methods of combining and when subjected to modeling of series with greater variability.

A realização de previsões de volatilidade é uma atividade de suma importância para empresas e agentes econômicos, entretanto utilizar-se de apenas um modelo para obtê-las pode não ser suficiente para incorporar todo o conhecimento associado ao ambiente de previsões. As técnicas de combinação de previsões podem incorporar todo o conhecimento associado ao ambiente de previsão. As técnicas de combinação têm como objetivo principal incorporar vários modelos com a finalidade de reduzir as medidas de erro de previsão. Este trabalho apresenta uma comparação da acurácia dos modelos individuais e das técnicas de combinação. Os modelos individuais incluídos nas técnicas de combinação são os modelos da família GARCH, o modelo de Alisamento Exponencial e o de Volatilidade Estocástica. Já as técnicas de combinação escolhidas foram a técnica de combinação por média aritmética, a técnica de combinação de pesos fixos proposta por Granger e Ramanathan (1984), a técnica de combinação com pesos móvel de Terui e Djik (2002). / The realization of forecasts of volatility is an activity of extreme importance for companies and economy agents, however to utilize only one model to obtain them could be insuficient to incorporate all the knowledge associated to the ambient of previsions. The technics of combination of forecasts have as its main objective to incorporate various models with the finality to reduce the measures of error of prediction. This work presents a comparision of the acuracy of the individual models and of the combination technics. The individual models included on the technics of combination are the models of the family GARCH, the model of Exponentially Weighted Moving Averages. Thus the technics of combination chosen were the technic of combination by arithmetic average, the technic of fixed weights proposed by Granger and Ramanathan (1984), the technic of combination of movable weights of Terui e Djik (2002).