Basketball field -goal percentage prediction model research and application based on BP neural network

摘要

Through ability analysis ofNBA teammembers, utilize BP neural network to make prediction on basketball field-goal percentage. Extract 50 feature vectors from 16 members??? stats to take training, it finds that outputs well conform to actual values, indicates that the received connection weight can reflect actual status. It mainly on the basis of time series statistical data, basketball competition statistical technique, back propagation neural network used model BP network, adopt rolling prediction to predict China athletes??? basketball field-goal percentage. Difference between prediction results and actual are relative small, it shows prediction feasibility of using BP neural network to define basketball field-goal percentage.