Adaptive Mixed-Attribute Data Clustering Method Based on Density Peaks

The clustering of mixed-attribute data is a vital and challenging issue. density peaks algorithm brings us simple efficient solution, but it mainly focuses on numerical attribute cannot be adaptive. In this paper, we studied the adaptive improvement method such an proposed based called AMDPC. algorithm, used unified distance metric to construct matrix, calculated local K-nearest neighbors, automatic determination cluster centers three inflection points. Experimental results real University California-Irvine (UCI) datasets showed that AMDPC could realize data, can automatically obtain correct number clusters, improved accuracy all by more than 22.58%, 24.25%, 28.03%, 22.5%, 10.12% for Heart, Cleveland, Credit, Acute, Adult compared traditional K-prototype respectively. It also outperformed modified (DPC_M) algorithms.