Although the Multi-Input and Multi-Output (MIMO) communication systems provide very high data rates with low error probabilities, these advantages are obtained at the expense of having high signal processing tasks and the hardware cost, e.g. expensive Analog-to-Digital (A/D) converters. The increased hardware cost is mainly due to having multiple Radio Frequency (RF) chains (one for each antenna element). Antenna selection techniques have been proposed to lower the number of RF chains and provide a low cost MIMO system. Among them, due to a beamforming capability Soft Antenna Selection (SAS) schemes have shown a great performance improvement against the traditional antenna sub-set selection methods for the MIMO communication systems with the same number of RF chains. A SAS method is basically realized by a pre-processing module which is located in RF domain of a MIMO system. In this thesis, we investigate on the receive SAS-MIMO, i.e. a MIMO system equipped with a SAS module at the receiver side, in noise-limited/interference channels. For a noise-limited channel, we study the SAS-MIMO system for when the SAS module is implemented before Low Noise Amplifier (LNA), so-called pre-LNA, under both spatial multiplexing and diversity transmission strategies. The pre-LNA SAS module only consists of passive elements. The optimality of the pre-LNA SAS method is investigated under two di erent practical cases of either the external or internal noise dominates. For the interference channel case, the post-LNA SAS scheme is optimized based on Power Angular Spectrum (PAS) of the received interference signals. The analytical derivations for both noise-limited and interference channels are verified via the computer simulations based on a general Rician statistical MIMO channel model. The simulation results reveal a superiority of the post-LNA SAS to the post-LNA SAS at any condition. Moreover, using the simulations performed for the interference channels we show that the post-LNA SAS is upper bounded by the full-complexity MIMO. Since in both above-mentioned channels, noise-limited and interference, the channel knowledge is needed for the SAS optimization, in this thesis we also propose a two-step channel estimation method for the SAS-MIMO. This channel estimation is based on an Orthogonal Frequency-Division Multiplexing (OFDM) MIMO system. Two di erent estimators of Least-Square (LS) and Minimum-Mean-Square- Error (MMSE) are applied. Simulation results show a superiority of the MMSE method to the LS estimator for a MIMO system simulated under the 802.16 framing strategy. Moreover, a 802.11a framing based SAS-MIMO is simulated using MATLAB SIMULINK to verify the two-step estimation procedure. Furthermore, we also employ a ray-tracing channel simulation to assess di erent SAS configurations, i.e. realized by active (post-LNA) and/or passive (pre-LNA) phased array, in terms of signal coverage. In this regard, a rigorous Signal to Noise Ratio (SNR) analysis is performed for each of these SAS realizations. The results show that although the SAS method performance is generally said to be upperbounded by a full-complexity MIMO, it shows a better signal coverage than the full-complexity MIMO.

Rapidly evolving Internet and web technologies and international efforts on standardization of learning object metadata enable learners in a web-based educational system ubiquitous access to multiple learning resources. It is becoming more necessary and possible to provide individualized help with selecting learning materials to make the most suitable choice among many alternatives.<p> A framework for individualized learning object selection, called Eliminating and Optimized Selection (EOS), is presented in this thesis. This framework contains a suggestion for extending learning object metadata specifications and presents an approach to selecting a short list of suitable learning objects appropriate for an individual learner in a particular learning context. The key features of the EOS approach are to evaluate the suitability of a learning object in its situated context and to refine the evaluation by using available historical usage information about the learning object. A Learning Preference Survey was conducted to discover and determine the relationships between the importance of learning object attributes and learner characteristics. Two weight models, a Bayesian Network Weight Model and a Naïve Bayes Model, were derived from the data collected in the survey. Given a particular learner, both of these models provide a set of personal weights for learning object features required by the individualized learning object selection.<p> The optimized selection approach was demonstrated and verified using simulated selections. Seventy simulated learning objects were evaluated for three simulated learners within simulated learning contexts. Both the Bayesian Network Weight Model and the Naïve Bayes Model were used in the selection of simulated learning objects. The results produced by the two algorithms were compared, and the two algorithms highly correlated each other in the domain where the testing was conducted.<p> A Learning Object Selection Study was performed to validate the learning object selection algorithms against human experts. By comparing machine selection and human experts selection, we found out that the agreement between machine selection and human experts selection is higher than agreement among the human experts alone.

Fall among the elderly is an important health issue. Fall detection and movement tracking techniques are therefore instrumental in dealing with this issue. This thesis responds to the challenge of classifying different movement types as a part of a system designed to fulfill the need for a wearable device to collect data for fall and near-fall analysis. Four different fall activities (forward, backward, left and right), three normal activities (standing, walking and lying down) and near-fall situations are identified and detected. Different machine learning algorithms are compared and the best one is used for the real time classification. The comparison is made using Waikato Environment for Knowledge Analysis or in short WEKA. The system also has the ability to adapt to different gaits of different people. A feature selection algorithm is also introduced to reduce the number of features required for the classification problem.

Given the breadth of constraint satisfaction problems (CSPs) and the wide variety of CSP solvers, it is often very difficult to determine a priori which solving method is best suited to a problem. This work explores the use of machine learning to predict which solving method will be most effective for a given problem. We use four different problem sets to determine the CSP attributes that can be used to determine which solving method should be applied. After choosing an appropriate set of attributes, we determine how well j48 decision trees can predict which solving method to apply. Furthermore, we take a cost sensitive approach such that problem instances where there is a great difference in runtime between algorithms are emphasized. We also attempt to use information gained on one class of problems to inform decisions about a second class of problems. Finally, we show that the additional costs of deciding which method to apply are outweighed by the time savings compared to applying the same solving method to all problem instances.

In this paper, we briefly describe the automatic model selection which is provided by Autometrics in the PcGive program. The modeler only needs to specify the initial model and the significance level at which to reduce the model. Then, the algorithm does the rest. The properties of Autometrics are discussed. We also explain its background concepts and try to see whether the model selected by the Autometrics can perform well. For a given data set, we use Autometrics to find a “new” model, and then compare the “new” model with a previously selected one by another modeler. It is an interesting issue to see whether Autometrics can also find models which fit better to the given data. As an illustration, we choose three examples. It is true that Autometrics is labor saving and always gives us a parsimonious model. It is really an invaluable instrument for social science. But, we still need more examples to strongly support the idea that Autometrics can find a model which fits the data better, just a few examples in this paper is far from enough.

Within any manufacturing environment, the selection of the production or assembly machines is part of the day to day responsibilities of management. This is especially true when there are multiple types of machines that can be used to perform each assembly or manufacturing process. As a result, it is critical to find the optimal way to select machines when there are multiple related assembly machines available. The objective of this research is to develop and present a model that can provide guidance to management when making machine selection decisions of parallel, non-identical, related electronics assembly machines. A model driven Decision Support System (DSS) is used to solve the problem with the emphasis in optimizing available resources, minimizing production disruption, thus minimizing cost. The variables that affect electronics product costs are considered in detail. The first part of the Decision Support System was developed using Microsoft Excel as an interactive tool. The second part was developed through mathematical modeling with AMPL9 mathematical programming language and the solver CPLEX90 as the optimization tools. The mathematical model minimizes total cost of all products using a similar logic as the shortest processing time (SPT) scheduling rule. This model balances machine workload up to an allowed imbalance factor. The model also considers the impact on the product cost when expediting production. Different scenarios were studied during the sensitivity analysis, including varying the amount of assembled products, the quantity of machines at each assembly process, the imbalance factor, and the coefficient of variation (CV) of the assembly processes. The results show that the higher the CV, the total cost of all products assembled increased due to the complexity of balancing machine workload for a large number of products. Also, when the number of machines increased, given a constant number of products, the total cost of all products assembled increased because it is more difficult to keep the machines balanced. Similar results were obtained when a tighter imbalance factor was used.

Water quality models are applied to assess the various processes affecting the concentrations of contaminants in a watershed. SPAtially Referenced Regression On Watershed attributes (SPARROW) is a nonlinear regression based approach to predict the fate and transport of contaminants in river basins. In this research SPARROW was applied to the Guadalupe and San Antonio River Basins of Texas to assess E. coli contamination. Since SPARROW relies on the measured records of concentrations of contaminants collected at monitoring stations for the prediction, the effect of the locations and selections of the monitoring stations was analyzed. The results of SPARROW application were studied in detail to evaluate the contribution from the statistically significant sources. For verification of SPARROW application, results were compared to 303 (d) list of Clean Water Act, 2000. Further, a methodology to maintain the monitoring records of the highly contaminated areas in the watersheds was explored with the application of the genetic algorithm. In this study, the importance of the available scale and details of explanatory variables (sources, land-water delivery and reservoir/ stream attenuation factors) in predicting the water quality processes were also analyzed. The effect of uncertainty in the monitored records on SPARROW application was discussed. The application of SPARROW and genetic algorithm were explored to design a monitoring network for the study area. The results of this study show that SPARROW model can be used successfully to predict the pathogen contamination of rivers. Also, SPARROW can be applied to design the monitoring network for the basins.

The ubiquity of circadian systems has allowed their characterization in a broad range of model systems, which has greatly improved knowledge of how these systems are organized and the vast range of cellular and organismal processes under circadian control. Most of the advances, however, have come in describing the central oscillators of these systems, and, in some cases, the input pathways used to coordinate these oscillators to external time. Very little progress has been made in understanding the output pathways that allow circadian systems to regulate the breadth of processes shown to be clock-controlled. A genetic selection was designed to obtain mutations in genes involved in circadianregulated expression of the Neurospora crassa ccg-1 and ccg-2 genes. Some, but not all, of the strains obtained display altered regulation of more than one ccg as well as an ‘Easlike’ appearance on solid media, and altered circadian period on race tubes. The data suggest a model in which output from the clock to these two genes is through a single, bifurcated pathway. The cloning of the gene mutated (rrg-1) in one of the strains from the above selection led to the first molecular description of a circadian output pathway in Neurospora, the HOG MAP kinase pathway. The HOG pathway has been previously described with regard to its role in the osmotic-stress response. The discovery of the involvement of rrg-1 in circadian regulation of ccg-1 and ccg-2 led to the discovery of regulation of the HOG pathway by the circadian clock. The data indicate that osmotic stress information and time-of-day information are transduced through the HOG pathway and implicate a role for the clock in preparing the organism for daily occurrences of hyperosmotic stress associated with sun exposure. The genetic selection, and the description of the HOG pathway with regard to circadian output, provide a basis for further characterization of circadian output in Neurospora. The ubiquity of MAP kinase pathways, such as the HOG pathway, and the observed similarities in the mechanisms of circadian clock function across multiple phyla, indicate that these findings may well be applicable to other model systems.

This study examined what effects, if any, cultural factors have on conceptual knowledge of evolutionary theory through natural selection. In particular, the study determines if Latino and non-Latino students differ in their misconceptions of natural selection and, if so, whether or not cultural factors could be the reason why such differences exist. A total of 1179 college students attending eight Hispanic-Serving Institutions in Texas participated in the study. The results revealed that the top two challenging natural selection concepts for students to comprehend were causes of phenotypic variation, i.e., mutations are intentional, and selective survival based on heritable traits. In addition, no statistical significant differences were found between the Latino and non-Latino students and the top four natural selection misconceptions between the groups were similar. Not even religion was found to directly contribute to evolutionary misconceptions; even though, it serves as the core of an individual’s beliefs system. However, traditional teaching methods, inadequately trained biology school teachers, lessons poor in content, insufficient teaching time, and lack of age appropriate tasks, as well as, poorly defined evolutionary terms are actually the main causes for evolutionary misconceptions.

Positive natural selection is recognized as the driving force underneath evolution. One of the surest signatures of recent positive selection is a local elevation of advantageous allele frequency and linkage disequilibrium (LD). This dissertation proposes a new test statistic to detect excess homozygosity based on a simple counting measure, which serves as a surrogate indicator of recent positive selection. Three tests are developed upon the new measure: (a) an extended genotype-based homozy- gosity test (EGHT), (b) a hidden Markov model test (HMMT), and (c) an extended haplotype-based homozygosity test (EHHT). The null hypotheses of all three tests assume random mating and Hardy-Weinberg equilibrium (HWE). They differ in how to treat LD under H0 . The EGHT assumes linkage equilibrium (LE) besides HWE while the EHHT allows arbitrary multi-locus LD. The HMMT stands between these two extremes and assumes pairwise but no higher-order disequilibrium interactions. We first conduct simulation study to compare the three score tests and verify that the EHHT is the most conservative one. We compare the performance of the EHHT with the prevailing detection methods and the EHHT has higher or similar power. We also evaluate the impact of simple demographic history on the EHHT and the simulation study suggests that the EHHT is resistant to the false-positive confounders resulting from simple demographic models. After extensive simulation studies, all three tests are then applied on HapMap Phase II data and we are able to replicate findings reported in the literature. We can also identify new candidate regions that may undergo recent selection through a set of filtering criteria including highest EHHT scores, high derived allele frequency and large population differentiation. Finally, we propose a cross-population comparison test statistic to detect chromosome regions in which there is no significant excess homozygosity in one population but homozygosity remains high in another population.