Evolutionary design of fuzzy classifiers using intersection points

Chromosome representation to search the optimal intersection points between adjacent fuzzy membership functions is originally presented for optimal design of fuzzy classifiers. Since the proposed representation contains the intersection points directly related to the boundary of classification, it is intuitively expected that redundancy of the search space is reduced and the performance is better in comparison with the conventional encoding scheme. Unlike the previous work, the distances between the intersection points are encoded instead of x-coordinates of intersection points in order to reduce the redundancy due to the combinations of disordered intersection points. The experimental results show that the proposed encoding scheme gives superior or competitive performance in two real-world datasets and gives more interpretable fuzzy classifiers. In addition, this proposed approach provides more interpretable classifiers without additional computational cost and also reduces search space while maintaining performance.