Research and Evaluation Based on Principal Component Analysis and BP Neural Networks

In this paper, based on the GBET dataset, 12 factors are used as secondary indicators, and then the 12 indicators are extracted into 3 primary indicators, and then the fuzzy comprehensive evaluation model is used for principal component analysis. Taking the 3 indicators as inputs and the evaluation results as outputs, the BP neural network model is used to establish the evaluation model. Considering the shortcomings of the BP neural network with low accuracy, we used a computer to analyze its error and improve the accuracy. Finally, we performed model testing to ensure the usability of the model.