Combining evolutionary, connectionist, and fuzzy classification algorithms for shape analysis

Abstract

This paper presents an investigation into the classification of a difficult data set containing large intra-class variability but low interclass variability. Standard classifiers are weak and fail to achieve satisfactory results however, it is proposed that a combination of such weak classifiers can improve overall performance. The paper also introduces a novel evolutionary approach to fuzzy rule generation for classification problems.