Exceptional rules are often ignored because of their small support. However, they have high confidence, so they are useful sometimes. A new algorithm for mining exceptional rules is presented, which creates a large it...Exceptional rules are often ignored because of their small support. However, they have high confidence, so they are useful sometimes. A new algorithm for mining exceptional rules is presented, which creates a large itemset from a relatively small database and scans the whole database only one time to generate all exceptional rules. This algorithm is proved to be quick and effective through its application in a mushroom database.展开更多
基金This project was supported by the National Natural Science Foundation of China (No. 69835010).
文摘Exceptional rules are often ignored because of their small support. However, they have high confidence, so they are useful sometimes. A new algorithm for mining exceptional rules is presented, which creates a large itemset from a relatively small database and scans the whole database only one time to generate all exceptional rules. This algorithm is proved to be quick and effective through its application in a mushroom database.
