A BP neural network model for intelligent quality monitoring of industry-education integration and talent cultivation

To achieve process-oriented and quantitative quality monitoring of the effectiveness of industry-education integration in universities, this study proposes a quality evaluation method based on the Back Propagation neural network (BPNN). The method conducts predictive modeling and performance verification for different types of university-enterprise cooperation projects in universities. Based on the Input-Process-Output (IPO) evaluation framework, this study constructs a multi-dimensional indicator system and develops an enhanced Back Propagation Neural Network (BPNN) model to improve its robustness and adaptability in heterogeneous educational data. The results show that the constructed BPNN achieves a mean squared error (MSE) of 0.0114 and a coefficient of determination (R²) of 0.947 on the validation set. Compared with the LR model, its R² is improved by 41.2%, ranking the best-performing model among all comparative models. Moreover, it maintains stable predictive performance with R² > 0.93 across eight professional subsets. The research indicates that the BPNN can effectively capture the complex relationships among multi-dimensional educational indicators and is suitable for quality evaluation tasks of industry-education integration in multiple scenarios of universities. This approach provides a robust, intelligent tool for enhancing the dynamic management and continuous improvement of industry-education integration initiatives in higher education.