摘要： 样本集中的数据数量过少会影响最后的分类精度。为此，提出基于元胞自动机模型的小样本集分类算法，利用元胞自动机的状态转换规则和分类规则，通过元胞空间的演化实现对数据集的分类。对标准数据集的测试结果表明，在小样本集和样本集中数据分布不均匀的情况下，该算法具有较高的分类精度。

关键词: 元胞自动机, 元胞邻域, 小样本集, 数据挖掘, 分类算法, 转换规则

Abstract: If the sample sets are very little, the accuracy of classification will be influenced. A classification algorithm based on Cellular Automata(CA) on small sample sets is presented to improve the classification accuracy by evolution of the cellular space, which uses the conversion rules of cellular automata and the rules of the classification. This algorithm has the characteristics of extending to high dimensional space and parallel computing, it can also classify data in the case of a small samples and uneven distribution of the data. Experimental results show that the algorithm can enhance the accuracy of classification.

Key words: Cellular Automata(CA), cellular neighbor, small sample set, data mining, classification algorithm, conversion rule

中图分类号: 

• TP301.6