In this thesis, we investigate partner selection schemes in multiuser cooperative networks. In networks, cooperative partners adopt coded cooperation to forward signals. Among the literature, two classes of two namely, centralized partner selections and distributed partner selections, have been proposed to select appropriate relays. Centralized partner selection is able to achieve the global optimization than distributed partner selection. However, centralized partner selections require high complexity, and global channel state information, which demands large amount of overhead and waste radio resources. Especially when the size of network increases, the cost to search appropriate relay for each user dramatically increases. Hence, we consider distributed partner selection scheme in the thesis. Among the existing work, fixed priority selection algorithm is a distributed partner selection algorithm strategy; where partner assignment is based on node indices do not include any channel information. To enhance performance, we exploit the geographical information of all users. Different from other distributed partner selection schemes, we adopt the method of Carrier Sense Multiple Access to exchange local information. We proposed coverage search algorithm, nearest source search algorithm and nearest middle-point search algorithm, the serve as the criteria of partner selection. The contributions of this thesis are to raise SNR, increase the probability that the achievable rate is great than the data rate in the distributed partner selection scheme, and enhance system performance.

The popularity of the World Wide Web (WWW) in providing a vast array of information has drawn a large number of users in the past few years. The dramatic increase in the number of Internet users, however, has brought undesirable impacts on users, such as long response time and service unavailability. The utilization of multiple servers can be used to reduce adverse impacts. The challenge is to identify a good resource site to allocate to the user given a group of servers from which to select. In this project, a performance-directed site selection system was developed for a web-based application called AADMLSS (African American Distributed Multiple Learning Styles System). Four different sets of experiments were conducted in this study. In order to evaluate the effectiveness of the test system, two other server selection methods, Load-based and Random-based methods, were implemented for comparative purposes. The experiments were also run during daytime and nighttime to see the impact of network load on the response time. Experimental results indicate that the performance-directed site selection system outperforms the Load-based and Random-based methods consistently. The response time is typically high during daytime and low during nighttime, indicating that the network load has an impact on the response time delivered. The results also show that server performance contributes to the overall response time, and network performance is the more dominating factor in determining a good resource site for the user.

The objective of the current study was to provide an initial empirical test of the Schneider, Smith, and Sipe (2000) multilevel model of personnel selection. The Schneider et al. (2000) model expanded the traditional approach to validating selection systems to include the impact that selection systems have on the broader organizational system. The current project provided an empirical test of this model by extending the traditional individual-differences approach to validation research and including group- and organization-criteria (e.g., unit-level performance and customer satisfaction). Using a quasi-experimental design, archival data from a managerial development and selection program were analyzed to examine several relationships proposed in the Schneider et al. (2000) model. The current study provided limited support for the Schneider et al. (2000) model. There were several limitations in the current study associated with the use of archival data, but the current study provides an initial indication of practical problems associated with empirically testing the model. While intuitively appealing, testing the Schneider et al. model in applied settings may prove to be a practical challenge because of the nature and complexity of the data required to do so. Although the current study provided limited support for the model, there were some interesting findings that warranted additional examination. Findings from the current study may be informative for both researchers and practitioners. Ideas for future research related to the Schneider et al. (2000) multilevel model of personnel selection are also offered.

Thesis (doctoral)--Universität, Freiburg (Breisgau), 2003. / Includes bibliographical references (p. [183]-200).

We perform a review of model selection procedures, in particular various cross validation procedures and adaptive model selection. We cover important results for these procedures and explore the connections between different procedures and information criteria.

The automated analysis of patients’ biomedical data can be used to derive diagnostic and prognostic inferences about the observed patients. Many noninvasive techniques for acquiring biomedical samples generate data that are characterized by a large number of distinct attributes (i.e. features) and a small number of observed patients (i.e. samples). Deriving reliable inferences, such as classifying a given patient as either cancerous or non-cancerous, using these biomedical data requires that the ratio r of the number of samples to the number of features be within the range 5 < r < 10. To satisfy this requirement, the original set of features in the biomedical datasets can be reduced to an ‘optimal’ subset of features that most discriminates the observed patients. Feature selection techniques strategically seek the ‘optimal’ subset. In this thesis, I present a new feature selection technique - multilevel feature selection. The technique seeks the ‘optimal’ feature subset in biomedical datasets using a multilevel search algorithm. This algorithm combines a hierarchical search framework with a search method. The framework, which provides the capability to easily adapt the technique to different forms of biomedical datasets, consists of increasingly coarse forms of the original feature set that are strategically and progressively explored by the search method. Tabu search (a search meta-heuristics) is the search method used in the multilevel feature selection technique. I evaluate the performance of the new technique, in terms of the solution quality, using experiments that compare the classification inferences derived from the result of the technique with those derived from the result of other feature selection techniques such as the basic tabu-search-based feature selection, sequential forward selection, and random feature selection. In the experiments, the same biomedical dataset is used and equivalent amount of computational resource is allocated to the evaluated techniques to provide a common basis for comparison. The empirical results show that the multilevel feature selection technique finds ‘optimal’ subsets that enable more accurate and stable classification than those selected using the other feature selection techniques. Also, a similar comparison of the new technique with a genetic algorithm feature selection technique that selects highly discriminatory regions of consecutive features shows that the multilevel technique finds subsets that enable more stable classification.

The automated analysis of patients’ biomedical data can be used to derive diagnostic and prognostic inferences about the observed patients. Many noninvasive techniques for acquiring biomedical samples generate data that are characterized by a large number of distinct attributes (i.e. features) and a small number of observed patients (i.e. samples). Deriving reliable inferences, such as classifying a given patient as either cancerous or non-cancerous, using these biomedical data requires that the ratio r of the number of samples to the number of features be within the range 5 < r < 10. To satisfy this requirement, the original set of features in the biomedical datasets can be reduced to an ‘optimal’ subset of features that most discriminates the observed patients. Feature selection techniques strategically seek the ‘optimal’ subset. In this thesis, I present a new feature selection technique - multilevel feature selection. The technique seeks the ‘optimal’ feature subset in biomedical datasets using a multilevel search algorithm. This algorithm combines a hierarchical search framework with a search method. The framework, which provides the capability to easily adapt the technique to different forms of biomedical datasets, consists of increasingly coarse forms of the original feature set that are strategically and progressively explored by the search method. Tabu search (a search meta-heuristics) is the search method used in the multilevel feature selection technique. I evaluate the performance of the new technique, in terms of the solution quality, using experiments that compare the classification inferences derived from the result of the technique with those derived from the result of other feature selection techniques such as the basic tabu-search-based feature selection, sequential forward selection, and random feature selection. In the experiments, the same biomedical dataset is used and equivalent amount of computational resource is allocated to the evaluated techniques to provide a common basis for comparison. The empirical results show that the multilevel feature selection technique finds ‘optimal’ subsets that enable more accurate and stable classification than those selected using the other feature selection techniques. Also, a similar comparison of the new technique with a genetic algorithm feature selection technique that selects highly discriminatory regions of consecutive features shows that the multilevel technique finds subsets that enable more stable classification.

Thesis (Ph. D.)--Oregon State University, 1993. / Typescript (photocopy). Includes bibliographical references (leaves 68-75). Also available on the World Wide Web.

Thesis (Ph. D.)--University of Texas at Austin, 1998. / Vita. Includes bibliographical references (leaves 157-172). Available also in a digital version from Dissertation Abstracts.

Thesis (doctoral)--University of Stockholm, 1984. / Cover title. Added t.p. laid in. Includes bibliographical references (leaves 24-26 (1st group)).