Mining Maximal Frequent Itemsets Based on Dynamic Ant Colony Optimization

Mining maximal frequent itemsets is to find maximal subsets that appear frequently in datasets, there were many algorithms to effectively solve MFI. Ant colony optimization (ACO) is a new method to solve MFI. However, there are two bottlenecks in which the ACO algorithm takes too much time and solves imprecisely for MFI. A dynamic ACO algorithm with Max-Min Ant System and association graph is proposed to mining maximal frequent itemsets. Firstly, Ant Colony road map is constructed, and then under the instruction of dynamic pheromone and heuristic to mining local maximal frequent itemsets, by way of new local and global update mechanism to mining global maximal frequent itemsets. Compared experiments show that this algorithm is fast and effective.