Possibilistic Causal Reasoning with Intensity Grades Using Conditional Causal Possibilities as Its Knowledge

Abstract

Many systems analyzed in the real world include events with two kinds of fuzziness-uncertainty and intensity degress. As for uncertainty, there is a long and rich history of studies using Probabilities, and some new studies employing Possibility theory, which is suitable for ordering based on uncertainty, have been published recently. Then, for intensity degrees of events, several causal reasoning methods, such as Fuzzy cognitive maps that conduct analysis through a simulation by forward reasoning using fuzzy causalities, Abductive reasoning that employs causal knowledge with degrees of occurrence in the back ward manner, and Intensity reasoning that uses the causal knowledge both in forward and backward manner, have been proposed. However, there are few studies that cover both uncertainty and intensity as well as both forward and backward causal reasoning. This paper proposes a new reasoning technique that can deal with all of them. The proposed approach is an expansion of a Possibilistic causal reasoning conducted in forward and backward manner using Conditional causal possibilities.