Data Mining Model Based on the Improved BP Neural Network Algorithm

Abstract

A data mining algorithm based on fuzzy FKM algorithm and BP was proposed to solve the problem of long training time and low efficiency when the sample data contains the attributes unrelated to target data. The attributes of in- put data was clustered by using FKM clustering algorithm, and the attributes with weak correlation to target data were abandoned, and then there remain the attributes with strong correlation to target data, which reduce the training samples of neural network, and improving the training efficiency of the network. Tests on forecasting the content of Hemoglobin in the body of children show that the proposed algorithm is very practicable and reliable.