Incremental clustering algorithm of mixed numerical and categorical data based on clustering ensemble

Traditional clustering methods have disadvantages of unsteadiness, randomness and low accuracy for classifying mixed numerical and categorical data. Therefore, the incremental clustering algorithm of mixed numerical and categorical data based on clustering ensemble is proposed, which adopts the results of several clustering to replace that of single clustering and modifies the design of threshold. The experiment results show that the improved algorithm has higher stability and accuracy by using the characters of existing data, and possess better effectiveness.