In this paper, we propose a statistics model to better understand the probabilistic interpretation for fuzzy sets. The model shows the relationship between the two kinds of statistics: the set-valued statistics for fuzzy sets and the conventional probability statistics for membership functions. Our model overcomes the shortcomings in previous approaches. The results of this paper provide a theoretical interpretation to Zadeh's insight that possibility relates to soft constraints restricting the values of language variables
Published in:
Fuzzy Systems, IEEE Transactions on
(Volume:4
,
Issue:
2
)
Date of Publication: May 1996