Outlier mining method for high dimensional massive data based on RBF neural network

Abstract

The density of high-dimensional massive data is relatively high, and it is difficult to mine outlier information. The traditional mining method is too slow to mine outliers in highdimensional massive data because the divided mining density is too balanced. Therefore, the research on the high-dimensional massive data outlier mining method based on RBF neural network is studied. In this method, RBF neural network structure and hidden layer parameters are set, and the mining density of highdimensional massive data is set to realize the outlier mining of high-dimensional massive data. The experimental results show that: compared with the traditional mining methods, the outlier mining method in this study can first complete the task of outlier mining in high-dimensional massive data. It can be seen that the mining method of this study improves the speed of data mining.