Using genetic algorithms to evolve a type-2 fuzzy logic system for predicting bankruptcy

Abstract

In this paper, we use GAs to design an interval type-2 fuzzy logic system (IT2FLS) for the purpose of predicting bankruptcy. The shape of type-2 membership functions, the parameters giving their spread and location in the fuzzy partitions and the set of fuzzy rules are evolved at the same time, by encoding all together into the chromosome representation. Type-2 FLSs have the potential of outperforming their type-1 FLSs counterparts, because a type-2 fuzzy set has a footprint of uncertainty that gives it more degrees of freedom. The enhanced Karnik-Mendel algorithms are employed for the centroid type-reduction and defuzzification stage. The performance in predicting bankruptcy is evaluated by multiple simulations, in terms of both in-sample learning and out-of sample generalization capability, using a type-1 FLS as a benchmark.