Rule Generating of Rough Sets Based on Bayesian Theory

The generating rule method is presented for incompatible and incomplete information of test data based on Bayesian theory. Firstly, the rule´s conditional probability is calculated when the certainty (reliability) of the test data is the prior probability and the samples (supportability) is posterior probability. Then, Those rules whose conditional probability is bigger than a given threshold value should be preserved. Lastly, the rule is generated by logic conjunction and disjunction of all the preserved rules. The example and application analysis indicate that the algorithm is clear, the calculating process is simple and it can be easily applied to computer programs, moreover, this method can avoid the knowledge distortion and the rule losing to the maximum for generating rule.