Rule Mining Algorithm with a New Ant Colony Optimization Algorithm

This work proposes an enhanced ant colony optimization algorithm for mining classification rule, called Threshold Ant Colony Optimization Miner (TACO-Miner). The main goal of TACO-Miner is to provide comprehensible classification rules which have higher predictive accuracy and simpler rule list. Experiments on data sets from UCI data repository are made to compare the performance of TACO-Miner with Ant-Miner, ACO-Miner and C4.5, a well known classification Algorithm. The results show that TACO-Miner performs better than Ant-Miner, ACO-Miner and C4.5 with respect to predictive accuracy and simplicity of the rule list mined with less computational complexity and costs.