Abstract
This paper considers a “body of evidence” on a finite set X, determined by a “basic probability assignment” (BPA) on X. This land of information on the set AT is ambiguous and so, in the applications, it is important to measure the degree of uncertainty involved in each BPA. In fact, several measures of uncertainty and entropy have been formulated and investigated in the setting of Dempster-Shafer Theory of Evidence. In this paper, we use the formal analogy underlying many of those measures of uncertainty. This moves us to define a family of fuzzy measures associated with each BPA. Thus, we propose a general formulation of the entropy of a BPA. Several known measures of entropy (including measures of imprecision and of randomness), as well as other new measures, are subsumed under our general scheme as particular cases. The basic properties of all those measures of entropy are also studied.