International journal of intelligent systems vol:11 issue:11 pages:879-891
In this article different aspects of the decision table formalism are discussed. First, crisp decision tables are defined and their construction is described. Next, fuzzy extensions are made to crisp decision tables in order to deal with imprecision and uncertainty. As a result, with crisp decision tables as special cases, a form of fuzzy decision tables is defined which include fuzziness in the conditions as well as in the actions. Consequently, the concept of completeness is introduced in the context of fuzzy decision tables. Furthermore, fuzzy consultation of decision tables is discussed, which allows decision making with fuzziness based on the matching between fuzzy conditions and the concept of fuzzy logical implication. (C) 1996 John Wiley & Sons, Inc.