| 摘要: |
| 针对传统的克隆选择算法可能存在的早熟收敛现象和缺少交叉操作问题，提出一种高效的克隆退火优化算法.该算法结合了模拟退火算法与免疫系统的克隆选择机制，并保持全局搜索和局部搜索的平衡，可以有效提高算法的搜索效率，从而加快算法的收敛速度.同时，提出一种品质因数模型来分析该算法的动态性能，并运用Markov链理论对其收敛性进行分析.最后，将该算法应用到关联规则数据挖掘中，取得了较为理想的实验结果. |
| 关键词: 品质因数 收敛性 马尔可夫链 克隆选择算法 群体多样性 |
| DOI：10.13328/j.cnki.jos.004911 |
| 分类号: |
| 基金项目:国家自然科学基金（61603420）；湖北省自然科学基金（2014CFB413）；中南民族大学中央高校基本科研业务费专项资金（CZY14007）；科学计算与智能信息处理广西高校重点实验室科学开放基金（GXSCIIP201412） |
|
| Optimization and Quality Factor of Clonal Selection Algorithm |
|
SHU Wan-Neng1, DING Li-Xin2
|
|
1.College of Computer Science, South-Central University for Nationalities, Wuhan 430074, China;2.Computer School, Wuhan University, Wuhan 430072, China
|
| Abstract: |
| To tackle the problem that traditional clonal selection algorithm may suffer from premature convergence phenomenon and is lack of crossover operator problems, this paper proposes a new efficient clonal annealing optimization algorithm. The proposed algorithm combines simulated annealing algorithm with clonal selection mechanism of immune system, and maintains the balance of global and local search. The algorithm can effectively improve search efficiency, so as to speed up the convergence rate. Meanwhile, a quality factor model is used to analyze the dynamic performance of the algorithm, and an analysis of its convergence is performed using Markov chain theory. Finally, the proposed algorithm is applied to the association rule data mining, achieving satisfactory results. |
| Key words: quality factor convergence Markov chain clonal selection algorithm population diversity |
