摘要
为了高效地从数据库中挖掘分类规则,提出一种将粒子群优化算法和遗传算法相结合的新算法。该算法的核心思想是对规则的前件进行固定长度编码,适应度函数的计算由分类规则的准确率、置信度、支持度和简洁度构成,从而实现基于两者混合算法的分类器设计。将该分类器与遗传算法分类器和粒子群算法分类器进行对比,实验结果表明,该分类器具有更高的分类准确率以及更快的收敛速度。
To efficiently mine the classification rule from database, a novel hybrid classification algorithm based on Particle Swarm Optilrdzation(PSO) and Genetic Algorithm(GA) is proposed. The core idea of the proposed algorithm is as follows: a new rule code with fixed length is proposed, a novel fitness function combined with accuracy, confidence, support and simplicity is constructed, and a hybrid heuristic classifier is accomplished. Experimental results show that the proposed classification algorithm acheives higher classification accuracy and lower running time compared with other classification algorithms.
出处
《计算机工程》
CAS
CSCD
北大核心
2008年第11期86-87,90,共3页
Computer Engineering
基金
河南省自然科学基金资助项目(0624010002)
郑州市科技攻关基金资助项目(2006-8-1)
关键词
数据挖掘
粒子群
遗传算法
分类器
分类规则
data inning
particle swarm
genetic algorithm
classifier
classification rule