A probabilistic interpretation of a case of inference involving fuzzy quantification

  • Authors:

  • Richard Willmott

  • Affiliations:

  • -

  • Venue:
  • Fuzzy Sets and Systems
  • Year:
  • 1985

Quantified Score

Hi-index 0.20

Visualization

Abstract

Many cases of approximate reasoning of the form ''XB are A, YA are C, so ZB and C'', where X, Y and Z are fuzzy quantifiers such as 'most', 'probably a few', etc., seem to involve a probabilistic component in the reasoning. To formalize such arguments this paper proposes a model in which fuzzy quantification is interpreted probabilistically, and the process of inference is carried out by a numerical computation, yielding a probability distribution which is approximated linguistically to give the conclusion. The process involves the use of the classical hypergeometric distribution.