该文提出了基于粗糙集的K类模式分类器的体系结构(RSPCMNNC),基于粗糙集理论提出了三个预处理算法,简化了分类器的结构,降低了学习难度,有效地避免产生过多的子网。样本空间基于最大均衡的策略来划分,保证BP算法在学习过程中的有效性。实验结果表明,该文提出的RSPCMNNC分类器显示出更高的识别率,对于实际应用中多特征模式的识别问题,具有很大的实用价值。 In this paper, the architecture of K-class pattern classifier based on rough set (RSPCMNNC) is proposed. Three preprocessing algorithms are proposed based on rough set theory, which simplifies the structure of classifier, reduces the learning difficulty and effectively avoids excessive generation Subnet. The sample space is divided based on the strategy of maximum equilibrium to ensure the validity of BP algorithm in the learning process. Experimental results show that the proposed RSPCMNNC classifier shows a higher recognition rate, which is of great practical value for the identification of multi-feature modes in practical applications.