摘要
Ant-Miner is an Ant Colony Optimization algorithm for classification task. This paper proposes an improved version of Ant-Miner, named mAnt-Miner , which is based on mAnt-Miner (Ant-Miner that uses a population of many ants). mAnt-Miner uses a simple and invariable heuristic strategy, that avoids it easily trapping in the local optimal solution and improves the efficiency of the algorithm. mAnt-Miner has been compared against Ant-Miner and mAnt-Miner in six public domain data sets. The results show that: 1) in term of predictive accuracy, mAnt-Miner is competitive with Ant-Miner and better than mAnt-Miner;2) mAnt-Miner is faster than Ant-Miner and mAnt-Miner;3) the difference of the rule simplicity between three algorithms is small.
- 出版日期2012
- 单位合肥工业大学