Discovering Fuzzy Censored Classification Rules (fccrs): a Genetic Algorithm Approach

Classification Rules (CRs) are often discovered in the form of ‘If-Then’ Production Rules (PRs). PRs, being high level symbolic rules, are comprehensible and easy to implement. However, they are not capable of dealing with cognitive uncertainties like vagueness and ambiguity imperative to real word decision making situations. Fuzzy Classification Rules (FCRs) based on fuzzy logic provide a framework for a flexible human like reasoning involving linguistic variables. Moreover, a classification system consisting of simple ‘If-Then’ rules is not competent in handling exceptional circumstances. In this paper, we propose a Genetic Algorithm approach to discover Fuzzy Censored Classification Rules (FCCRs). A FCCR is a Fuzzy Classification Rule (FCRs) augmented with censors. Here, censors are exceptional conditions in which the behaviour of a rule gets modified. The proposed algorithm works in two phases. In the first phase, the Genetic Algorithm discovers Fuzzy Classification Rules. Subsequently, these Fuzzy Classification Rules are mutated to produce FCCRs in the second phase. The appropriate encoding scheme, fitness function and genetic operators are designed for the discovery of FCCRs. The proposed approach for discovering FCCRs is then illustrated on a synthetic dataset.