Belief Functions: A Revision of Plausibility Conflict and Pignistic Conflict

Authors: Milan Daniel
Publication year: 2013
Publication type: Konferenční příspěvek (mezinárodní konf.)
Periodical: Scalable Uncertainty Management
Projects: Mathematical Fuzzy Logic in Computer Science
Bibliographic citation: DANIEL, Milan. Belief Functions: A Revision of Plausibility Conflict and Pignistic Conflict. In: LIU, W., SUBRAHMANIAN, V.S., WIJSEN, J., eds. Scalable Uncertainty Management. Berlin: Springer, 2013, S. 190-203. Lecture Notes in Artificial Intelligence, 8078. ISBN 978-3-642-40380-4. ISSN 0302-9743
Dostupné z: doi:10.1007/978-3-642-40381-1_15
DOI: 10.1007/978-3-642-40381-1_15
ID: 396394

Abstract

Plausibility conflict of belief functions is based on decisional support / opposition of elements of a frame of discernment. It distinguishes conflict between belief functions from internal conflicts of individual functions. This contribution presents a revision of plausibility conflict between belief functions. According to four types of conflicting sets, four variants of plausibility conflict are defined. Further, a new alternative approach - pignistic conflict - based on pignistic probability instead of on normalized plausibility of singletons is introduced. Its cautious version may be considered to be an improvement of Liu’s degree of conflict cf. Comparing the approaches, a relation of sum of conflicting belief masses and a relation of a distance of belief functions to conflict between belief functions are also discussed.