When In Doubt,BeIndecisive
Stc
Date: 2010-12-03
Time: 11:00
Room: Minnaert 211
Speaker: Linda van der Gaag
Title: When in Doubt ... Be Indecisive
Abstract
For a presented case, a Bayesian network classifier in essence computes a
posterior probability distribution over its class variable. Based upon this
distribution, the classifier's classification function returns a single,
determinate class value and thereby hides the uncertainty involved. To provide
reliable decision support, however, the classifier should be able to convey
indecisiveness if the posterior distribution computed for the case does not
clearly favour one class value over another. In this talk, an approach for
this purpose is presented, and new measures are introduced to capture the
performance and practicability of such classifiers. The talk is based upon
joint work with Silja Renooij, Wilma Steeneveld and Henk Hogeveen.
