针对离散制造业质量管理系统中维度高,且存在较多一致性数据的情形,设计了一种基于信息论中的信息熵,即互信息的改进聚类算法。通过实验分析,采用改进的聚类算法可有效提高聚类的正确率,并且通过演化聚类理论和方法的应用可对离散制造业质量管理提供有效的决策参考信息。
For high dimension and much consistent data in discrete manufacturing quality management system, an improved clustering algorithm is designed based on information entropy in information theory which is also called mutual information. Experimental analysis shows that using the improved clustering algorithm can effectively improve the correct rate of clustering, and the application of the evolution of the cluster theory and method can provide effective decision-making reference information for discrete manufacturing quality management.