none

<p>This work focus on situation prediction in data fusion systems.</p><p>A hypothesis evaluation algorithm based on artificial neural networks is introduced. It is evaluated and compared to an algorithm based on Bayesian networks which is commonly used. It is also compared to a simple "dummy" algorithm.</p><p>For the tests, a computer based model of the environment, including protected objects and enemy objects, is implemented. The model handles the navigation of the enemy objects and situational data is extracted from the environment and provided for the hypothesis evaluation algorithms.</p><p>It was the belief of the author that ANNs would be suitable for hypothesis evaluation if a suitable data representation of the environment were used. The representation requirements include pre processing of the situational data to eliminate the need for variable input size to the algorithm. This because ANNs poorly handles this; the whole network have to be retrained each time the amount of input data changes.</p><p>The results show that ANNs performed best of the three and hence seems to be suitable for hypothesis evaluation.</p>

No description available.

Thesis (Ph. D.)--University of Texas at Austin, 2000. / Vita. Includes bibliographical references (leaves 149-160). Available also in a digital version from Dissertation Abstracts.

Research on exposure to animals and risk of type 1 diabetes (T1D) has had conflicting results with some researchers finding that animal exposure reduces the risk of T1D and others finding no association between animal exposure and T1D. Previously conducted studies on the association between animal exposure and T1D are case-control studies that have been limited by recall bias. The purpose of this study is to investigate the association between early life animal exposure and the risk of persistent, confirmed islet autoantibodies (IA) and T1D diagnosis among an eligible cohort of genetically high T1D risk participants enrolled in the international prospective cohort study, The Environmental Determinants of Diabetes in the Young (TEDDY). It is hypothesized that children who are exposed to animals in early life will have a lower risk of developing IA and T1D than children who are not exposed to animals in early life. A total of 7,432 TEDDY participants were included in the study. The associations between early life animal exposure and the outcomes of interest were explored using Cox proportional hazards models. In order to control for confounding, a propensity score analysis was applied by three different methods: adjustment for the propensity score variable in the Cox proportional hazards model, stratification on propensity score groups, and propensity score pair matching. Early life animal exposure was not associated with diabetes autoimmunity or T1D onset in this genetically high T1D risk population. These findings were consistent across all three propensity score analysis methods and when directly adjusting for HLA type. The hypothesis that children who are exposed to animals in early life will have a lower risk of developing IA and T1D than children who are not exposed to animals in early life is not supported by this study. The results of this study suggest that there is no association between early life animal exposure and development of T1D. Performing this analysis again after longer follow-up has been completed for the study population is recommended as it may elucidate the effect of animal exposure on T1D and IA risk. Further studies are also needed on animal exposure and T1D in different types of environments (e.g., high residential density) and the effect of different types of animal exposures (e.g., species, duration) on T1D and IA risk. Additionally, studies on differences in perceptions of pets across countries could also aid the interpretation of studies on animal exposure and health outcomes.

According to the Body-Specificity Hypothesis, experiences of habitual motor fluency cause people to associate positive valence with their dominant hand side and confer positive valence to items located on their dominant hand side (Casasanto, 2009). Can ongoing motor experience impact this association in the absence of visually lateralized stimuli? In Experiment 1, participants flipped cards using one hand and rated the image on each card with respect to how well it was described by positive or negative personal characteristics. Contrary to our predictions, participant’s ratings were not biased by the hand that they used during the trial. In Experiment 2, the task was almost entirely the same, though participants wore a slippery glove on their dominant hand to reduce the perceived motor fluency of the dominant hand. Again, participant’s ratings were not biased by the relative motor fluency of the hand used during the trial. Results indicate that ongoing motor activity may not be sufficient to activate body specific preferences in the absence of visually lateralized stimuli. / text

The aim of this research is to investigate the long-run co-integrating relationships in the Asian markets. Our research focuses on 4 areas; pair trading, out-of-sample forecasting, testing the unbiased forward exchange rate hypothesis and testing the expectation hypothesis of the term structure of interest rates. The introduction is provided in chapter one. In chapter two, we develop a pairs trading strategy using individual stocks listed in the Stock Exchange of Thailand. Engle and Granger approach is used to identify the potential pairs that are cointegrated. The results show that pairs trading strategy is profitable in this market. Chapter three examines the forecasting performance of the error correction model on daily share price series from the Stock Exchange of Thailand. The disequilibrium term is classified into “correct” and “mix” sign based on Alexander (2008)’s criterion; the results indicate that the error correction component can help to improve the predictability in the long run. Chapter four tests the unbiased forward rate hypothesis of 11 Asian exchange rates using linear conventional regression, ECM and logistic smooth transition regression with the forward premium as the transition variable. Out-of-sample forecasting results also suggest that inferior forecasting performance could be obtained as a result of using linear models. In chapter five, we investigate the expectation hypothesis of the term structure of interest rate for four Asian countries. We employ linear models and nonlinear approaches that allow to capture asymmetric and symmetric adjustments. The result also indicates that the term structure can be better modeled by means of LSTR models. The forecasting exercise also confirms these findings.

Phytophagous insects and plants comprise one half of all macroscopic organisms on Earth, and understanding plant-insect interactions has been a long-standing focus in ecological and evolutionary biology. A key challenge has been determining the factors that affect how insects distribute among plants. While ditrophic and tritrophic interactions are well known and enormously important, the biophysical context in which these interactions occur is largely unexplored. This dissertation examines how a plant's physical environment affects insect performance on, and preference for, its leaves.An insect's primary physiological challenges are staying within an appropriate temperature range and retaining sufficient water. These problems are exacerbated during the egg stage. Eggs have comparatively enormous ratios of surface area to volume, and their temperature is determined largely by where they are laid. Because they are small, eggs are nearly always immersed within their plant's boundary layer, a thin layer of still air that resists heat and moisture transfer between the plant and its surroundings. Almost no work has documented the microclimate to which insects are exposed in a plant's boundary layer, which likely differs substantially both from the ambient macroclimate, and from leaf to leaf.Because a female controls the location in which her eggs must develop, her choice of oviposition site may profoundly influence the success of her offspring. In this dissertation I examine how site-specific environmental variables, including microclimate, predation, and leaf nutrition, drive female oviposition preference and offspring performance in the hawkmoth, Manduca sexta L. (Lepidoptera: Sphingidae). I measure how leaves of M. sexta's primary host plants in the southwestern USA modify the temperature and humidity experienced by eggs, and how these leaf microclimates affect the performance of eggs and larvae. I then test whether this species' oviposition-site choices correlate with offspring performance with regard to microclimate, predation risk, and leaf nutrition. This dissertation is unique in focusing on the relatively unstudied biophysical context in which plant-insect interactions occur. Additionally, it is the first work that compares, together in a single study, the effects of varying multiple factors related to oviposition-site choice across all life history stages in a single model system.

No description available.

No description available.