Abstract: This article introduces a kind of classification algorithm based on fuzzy rule and genetic algorithm:firstly,provide one fuzzy rule generating method;secondly,genetic algorithm optimizes choosing the fuzzy rule;finally,start the trigger to the sample that can't be discerned.This algorithm can improve the recognition capability of samples and reduce the rule sets.The performance of the classification algorithm is tested by the IRIS data,and the results show that not only it has very good classification performance but also the fuzzy rules sets is simplified.

Key words: classification algorithm, classification correctness, rules simplify, genetic algorithm