Thesis (Ph. D.)--University of Maryland, College Park, 2005. / Includes bibliographical references (p. 167-176). Also available online via the University of Maryland digital repository website (https://drum.umd.edu/).

Project (M.S.)--Brigham Young University. Dept. of Statistics, 2005. / Includes bibliographical references (p. 51-52).

Thesis (Ph. D.)--Ohio State University, 2006. / Available online via OhioLINK's ETD Center; full text release delayed at author's request until 2007 Aug 14

Thesis (Ph. D.)--Ohio State University, 2006. / Title from first page of PDF file. Includes bibliographical references (p. 67-72).

Thesis (Ph. D.)--Drexel University, 2007. / Includes abstract and vita. Includes bibliographical references (leaves 103-115).

The problem of handling dependent evidence is an important practical issue for applications of reasoning with uncertainty in artificial intelligence. The existing solutions to the problem are not satisfactory because of their ad hoc nature, complexities, or limitations. In this dissertation, we develop a general framework that can be used for extending the leading uncertainty calculi to allow the combining of dependent evidence. The leading calculi are the Shafer Theory of Evidence and Odds-likelihood-ratio formulation of Bayes Theory. This framework overcomes some of the disadvantages of existing approaches. Dependence among evidence from dependent sources is assigned dependence parameters which weight the shared portion of evidence. This view of dependence leads to a Decomposition-Combination method for combining bodies of dependent evidence. Two algorithms based on this method, one for merging, the other for pooling a sequence of dependent evidence, are developed. An experiment in soybean disease diagnosis is described for demonstrating the correctness and applicability of these methods in a domain of the real world application. As a potential application of these methods, a model of an automatic decision maker for distributed multi-expert systems is proposed. This model is a solution to the difficult problem of non-independence of experts. / Graduation date: 1991

In Bayesian analysis, means are commonly used to summarize Bayesian posterior distributions. Problems with a large number of parameters often require numerical integrations over many dimensions to obtain means. In this dissertation, posterior modes with respect to appropriate measures are used to summarize Bayesian posterior distributions, using the Newton-Raphson method to locate modes. Further inference of modes relies on the normal approximation, using asymptotic multivariate normal distributions to approximate posterior distributions. These techniques are applied to two statistical estimation problems. First, Bayesian sequential dose selection procedures are developed for Bioassay problems using Ramsey's prior [28]. Two adaptive designs for Bayesian sequential dose selection and estimation of the potency curve are given. The relative efficiency is used to compare the adaptive methods with other non-Bayesian methods (Spearman-Karber, up-and-down, and Robbins-Monro) for estimating the ED50 . Second, posterior distributions of the order of an autoregressive (AR) model are determined following Robb's method (1980). Wolfer's sunspot data is used as an example to compare the estimating results with FPE, AIC, BIC, and CIC methods. Both Robb's method and the normal approximation for estimation of the order have full posterior results. / Graduation date: 1992

University of Macau / Faculty of Science and Technology / Department of Mathematics

Thesis (Ph.D.)--Utrecht University, 2006. / Includes bibliographical references (p. [133]-137).

Thesis (Ph. D.)--University of Hong Kong, 2001. / Includes bibliographical references (leaves 192-202).