摘要
数据分类是数据挖掘中的一个重要课题,研究各种高效的分类算法是数据挖掘的重要问题之一.本文对了GAAA算法进行改进提出了一种新组合优化算法,将其应用到分类规则的优化问题,采用遗传算法生成信息素分布,利用蚂蚁算法求精确解,优势互补,有效地节省了计算时间,并优化了生成的分类规则.实验结果表明:该算法可以有效克服停滞,提高搜索效率,有效地挖掘出最优的分类规则集.
Data classification of data mining is an important subject, how to study various efficient data mining classification algorithm is one of the important issues. This article improved GAAA algorithm and make a new portfolio optimization, it~ applied to the optimization problem of classification rules. Genetic algorithm generate distribution of information factors, ant algorithms have exact solutions, advantages Complement each other, effectively saving computing time, and to optimize the generation of classification rules. The results showed that: the algorithm can effectively overcome the stagnation, improve the efficiency of search and effectively mine the optimal set of classification rules.
出处
《辽宁大学学报(自然科学版)》
CAS
2010年第1期40-44,共5页
Journal of Liaoning University:Natural Sciences Edition
关键词
决策树
蚁群算法
遗传算法
decision tree
ant algorithm
genetic algorithm