Implementing a bayesian scheme for revising belief commitments : Lashon B. Booker, Naveen Hota, and Gavin Hemphill Navy Center for Applied Research in AI, Code 5510, Naval Research Laboratory, Washington, DC 20375

Abstract

Our previous work on classifying complex ship images [1,2] has evolved into an effort to develop software tools for building and solving generic classification problems. Managing the uncertainty associated with feature data and other evidence is an important issue in this endeavor. Pearl [7-11] has developed a Bayesian framework for managing uncertainty that has proven to be applicable to several of the belief maintenance functions that are necessary for classification problem solving. One such function is to determine a belief commitment which designates in categorical terms the most probable instantiation of all hypothesis variables given the evidence available. Before these belief commitments can be computed, the straightforward implementation of Pearl's procedure involves finding an analytical solution to some often very difficult optimization problems. We describe an implementation of this procedure using tensor products that solves these problems enumeratively and avoids the need for case by case analysis. The procedure is thereby made more practical to use in the general case.