Abstract
This paper focuses on research of an adaptive clustering algorithm and its application in medical diagnosis based on probabilistic neural networks. A PNN-CAdaBoost medical diagnosis model is proposed on the standard AdaBoost algorithm together with clustering algorithm. Both PNN-AdaBoost model and PNN-CAdaBoost model are established respectively. For testing our model validness, the experimental data are collected from the Wisconsin breast cancer data set in the UCI database, computations and comparisons with multiple indicators. It is proved that the PNN-CAdaBoost medical diagnosis model can effectively improve the classification performance and has good robust stability.