http://hdl.handle.net/1808/133 Search the Channel search http://kuscholarworks.ku.edu/dspace/simple-search http://hdl.handle.net/1808/467 Title: Inference in Hybrid Bayesian Networks with Mixtures of Truncated Exponentials<br/><br/>Authors: Cobb, Barry R.; Shenoy, Prakash P.<br/><br/>Abstract: Mixtures of truncated exponentials (MTE) potentials are an alternative to discretization for solving hybrid Bayesian networks. Any probability density function (PDF) can be approximated with an MTE potential, which can always be marginalized in closed form. This allows propagation to be done exactly using the Shenoy-Shafer architecture for computing marginals, with no restrictions on the construction of a join tree. This paper presents MTE potentials that approximate an arbitrary normal PDF with any mean and a positive variance. The properties of these MTE potentials are presented, along with examples that demonstrate their use in solving hybrid Bayesian networks. Assuming that the joint density exists, MTE potentials can be used for inference in hybrid Bayesian networks that do not fit the restrictive assumptions of the conditional linear Gaussian (CLG) model, such as networks containing discrete nodes with continuous parents.<br/><br/>Description: Has been accepted for publication in the International Journal of Approximate Reasoning, Elsevier Science Publishing Co., Inc. Wed, 01 Jun 2005 00:00:00 GMT http://hdl.handle.net/1808/159 Title: Sequential Valuation Networks for Asymmetric Decision Problems<br/><br/>Authors: Demirer, Riza; Shenoy, Prakash P.<br/><br/>Abstract: This paper deals with representation and solution of asymmetric decision problems. We describe a new representation called sequential valuation networks that is a hybrid of Covaliu and Oliver’s sequential decision diagrams and Shenoy’s valuation networks. The solution algorithm is based on the idea of decomposing a large asymmetric problem into smaller sub-problems and then using the fusion algorithm of valuation networks to solve the sub-problems. Sequential valuation networks inherit many of the strengths of sequential decision diagrams and valuation networks while overcoming many of theirshortcomings. We illustrate our technique by representing and solving a modified versionof Covaliu and Oliver’s Reactor problem in complete detail.<br/><br/>Description: This paper is currently (August 2004) in press at the European Journal of Operational Research. Mon, 01 Jan 2001 00:00:00 GMT http://hdl.handle.net/1808/156 Title: On Transforming Belief Function Models to Probability Models<br/><br/>Authors: Cobb, Barry R.; Shenoy, Prakash P.<br/><br/>Abstract: In this paper, we explore methods for transforming a belief function model to an equivalent probability model. We propose and define the properties of a method called the plausibility transformation method. We compare the plausibilitytransformation method with the pignistic transformation method. These two methods yield qualitatively different probability models. We argue that the plausibility transformation method is the correct method that maintains belief function semantics.<br/><br/>Description: In response to reviewer comments on this paper, we have written a shorter and more focused paper: "On the Plausibility Transformation Method for Translating Belief Function Models to Probability Models," University of Kansas School of Business Working Paper No. 308, June 2004, Lawrence, KS. Tue, 01 Jul 2003 00:00:00 GMT http://hdl.handle.net/1808/143 Title: Local Computation in Hypertrees<br/><br/>Authors: Shafer, Glenn; Shenoy, Prakash P.<br/><br/>Abstract: The monograph describes theory and algorithms for computation of marginals using local computation that applies to a large number of domains including probability theory, Dempster-Shafer theory of belief functions, discrete optimization, and constraint satisfaction.<br/><br/>Description: This is an unpublished monograph that was widely distributed (and cited). It was first written in August 1988 and subseqently revised. Mon, 01 Jul 1991 00:00:00 GMT