Abstract: This paper has proposed a fast algorithm for mining association rules based on half-spaces and GA. Most traditional algorithms are constrained by many aspects of practice, such as data type, real-life significance and so on, which probably constrain the capacity of knowledge acquisition seriously. By contraries, the algorithm proposed could get rid of the constraints. Moreover, besides supplying some interesting rules to user, it also does well in mining for large data set.

Key words: half-spaces, GA, association rules, data mining

摘要： 本文提出了一种利用半空间模型和遗传算法(GA)对关联规则进行快速挖掘的方法。传统关联规则挖掘算法往往受到数据类型、关联规则的实际意义等约束，大大限制了知识获取的能力。而本文提出的方法不再受到上述限制的困扰，并且可以挖掘出用户感兴趣的规则，尤其对于大规模样本集的效果也是相当不错的。

关键词: 半空间, 遗传算法, 关联规则, 数据挖掘