An Adaptive Ant Colony Algorithm for Classification Rule Mining

DOI

10.2991/aiie-16.2016.68How to use a DOI?

Keywords

data mining; adaptive ant colony algorithm; classification rule; pheromone

Abstract

Ant-Miner algorithm is a typical classification rule mining algorithm which can improve the classification accuracy and generate simple rules. However, it also has a few disadvantages, such as complicated computing method for heuristic factor, long calculation time, slow evolution and so on. Based on the Ant-Miner algorithm, this paper adjusted the probability of the deterministic selection and the volatility coefficient dynamically and adaptively. This not only guarantees the convergence speed but also improves the global search ability. To verify the effectiveness of the algorithm, we used public database UCI datasets for algorithm simulation. Compared with the Ant-Miner algorithm, the proposed algorithm improves the classification accuracy rate and gives more concise rules.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Xiaomeng Zhang
AU  - Wensheng Sun
PY  - 2016/11
DA  - 2016/11
TI  - An Adaptive Ant Colony Algorithm for Classification Rule Mining
BT  - Proceedings of the 2016 2nd International Conference on Artificial Intelligence and Industrial Engineering (AIIE 2016)
PB  - Atlantis Press
SP  - 295
EP  - 299
SN  - 1951-6851
UR  - https://doi.org/10.2991/aiie-16.2016.68
DO  - 10.2991/aiie-16.2016.68
ID  - Zhang2016/11
ER  -