Research And Application Of The Heuristic Attribute Reduction In Rough Set

Along with the extensive usage of the large-scale database and the fiercely fast expansion of Internet, the data saved in the database is quickly enlarging in the global. People urgently hope that they can dig out the important information in the amount of data.Rough Set is a kind of mathematics theories, which handles inaccurate and incomplete data. It is very important to the artificial intelligence and the cognition science. And it has already been extensively applied to the artificial intelligence, pattern recognition and data mining, etc. The theory has already been confirmed that it is very useful in practice. -Attribute reduction is a significant topic of rough set theory. Under the condition of unchanged classification and decision abilities, attribute reduction is to delete irrelative or unimportant attribute. People always expect to find out minimum reduction, but it has already been proved that it is a NP complete problem.In my paper I briefly introduce several kinds of heuristic attribute reduction algorithm, including heuristic attribute reduction algorithm which is based on discernibility, discernible matrix, information entropy, relative different table and genetic algorithm. Put forward two kinds of algorithms on this foundation, namely a kind of based on discernibility heuristic attribute reduction modified algorithm and a kind of heuristic attribute reduction algorithm based on genetic algorithm. Through the analysis of these two kinds of algorithm in examples and the application in the UCI data concentration, they are proved to be feasible. Moreover, the satisfied result for most people can be gotten.