Effect ofNumber ofTraining Samples on ANN Prediction Accuracy for Cutting Force

In order to predictcutting force using training samples as few as possible, training samples with differentnumberswere selected to train an artificialneuralnetwork (ANN) respectively, and the effectof the numberof training samples onANN prediction accuracy for cutting force based on the LM (Lenvenberg-Marquardt) algorithm was analyzed by contrast experiments. Statistic mean amplitude and mean square errorwere taken as the evaluation indexes for forecast results, and the relationship between the prediction accuracy for cutting force and the numberof training sampleswas investigated. The research result indicates that 40 to 50 groups of training samples may be sufficient to obtain accurate cutting forcewithin the certain range of cutting parameters.