New knowledge acquisition method in incomplete information system based on rough set and self-adaptive genetic algorithm

In this paper, a knowledge acquisition method is proposed in incomplete information system. Firstly, according to the rough set theory some improvement is made to the discernible matrix; then with integration of the self-adaptive genetic algorithm, a new knowledge acquisition algorithm to mine the consistent rules and inconsistent rules from the incomplete information system is presented. Furthermore, according to the study of confidence degree and support degree of association rule in complete information system, the rule confidence degree and support degree is defined within the framework of incomplete information system. Through the examination of experimental results, this algorithm is proved effective, especially to large incomplete information systems.