Models for computer reasoning under uncertainty

CINF 53

Philip N. Judson, LHASA Ltd, Department of Chemistry, University of Leeds, Leeds, LS2 9JT, United Kingdom
Human decision makers appear to reach their conclusions through reasoning based on weighing the arguments for and against propositions. In spite of their fallibilities, humans perform well enough by this means for evolutionary success and so their methods deserve consideration as models for computer reasoning. Bulding on earlier work on the logic of argumentation, at LHASA we have developed models for reasoning about the likelihood that an event or circumstance will come about and about whether some events are more or less likely than others. This talk will outline the background to our work and present some of the key features of the models we have developed. We use them in programs for predicting toxicity and xenobiotic metabolism but they are suitable for any area where predictions have to be based on uncertain information.
 

Herman Skolnik Award Symposium
8:15 AM-11:45 AM, Tuesday, August 24, 2004 Pennsylvania Convention Center -- 110A&B, Oral

Division of Chemical Information

The 228th ACS National Meeting, in Philadelphia, PA, August 22-26, 2004