Study on Knowledge Discovery for Lifestyle Diseases Using Rough Set

Because of the complexity and incompletion of medical data, as well as the rough set can be well to deal with fuzzy and uncertain information, a knowledge discovery model is presented for lifestyle diseases based on rough set theory. As the problem that single attribute dependency is difficult to distinguish, an improved heuristic algorithm is presented for attribute reduction based on the core attribute sets dependency. The dependency of core and other attributes are calculated in this paper to determine the dependency degree, so as to thoroughly study the minimum attribute reduction. Decision table and new rules can be established according to the reduced attributes, then the effective rules are extracted by higher accuracy and support. According to the reduced attributes to establish decision table and generate new rules, the effective rules are extracted by higher accuracy and support. Additionally, the process of constructing knowledge base is verified by clinical experiment. The results show that presented method can not only calculate the effective reduction, but also guarantee the high decision accuracy. Meanwhile, rules consistent with expert knowledge can be produced.