In view of the problem of CNC milling machine energy efficiency prediction caused by multiple influencing factors and complex correlation between factors,a convolution neural network (CNN) based method for CNC milling machine energy efficiency level prediction was proposed.Through the analysis of the influencing factors of energy efficiency in the operational process of CNC machine,the influencing factors were classified from the dimensions of equipment,process,workpiece,tool.According to the sources of different dimension data,the method of collecting and preprocessing multidimensional data of CNC milling machine was put forward.An energy efficiency level prediction method based on LeNet-5 improved convolutional neural network was proposed.The feasibility and applicability of the method were verified by a case.The final training accuracy reaches 97.29%,and the accuracy of the test set reaches 93.32%,with better prediction results.This method can be used to guide the selection of equipment and controllable parameters and has a good application prospect.