Geometric compatibility modification in FuzzyCLIPS

Abstract

The fuzzy inference method in most of the marketed development tools for fuzzy systems define rule relations using the min operator and employ the compositional rule of inference. This method is equivalent to set-theoretic compatibility modification inference Geometric compatibility modification inference, however, employs the dissemblance compatibility measure from metric class of compatibility measures. This inference method can determine a conclusion or an action even when the fuzzy input does not intersect the fuzzy antecedent of any rule such as in sparse rule bases. No set-theoretic compatibility measure has this capability. Because of this capability, the dissemblance compatibility measure is currently being implemented and tested as a new inference type in FuzzyCLIPS, a fuzzy systems development tool available on the Internet from the National Research Council of Canada.