A multi-level ant-based algorithm for fuzzy data mining
Tzung‐Pei Hong, Ya-Fang Tung, Shyue-Liang Wang, Yu-Lung Wu · 2009
In the past, we proposed a mining algorithm to find suitable membership functions for fuzzy association rules based on the ant colony systems. In that approach, the precision was limited since binary bits were adopted to encode the membership functions. The paper thus extends the original approach for increasing the accuracy of the results by adding multi-level processing. The membership functions derived in a level will be refined in the next level. The final membership functions in the last level are then output to the rule-mining phase for finding fuzzy association rules.