Possibility expectation and its decision making algorithm
James M. Keller, Bolin Yan · 2003
Both the theory and a decision making algorithm for a variation of the fuzzy integral are presented. This integral is based on a possibility measure where it is not required that the measure of the universe by unity. A training algorithm for the possibility densities in a pattern recognition application is also presented. The algorithm was run on the IRIS data set and the results compared to results for the Bayesian classifier.>