Teaching quality evaluation model construction of university industry and education integration by adaptive BP and DDAE-SVR

Abstract To adapt to changing teaching modes in colleges and universities, and better predict and analyze the students' learning situation, this study aims at a series of deficiencies existing in the traditional teaching quality evaluation index system. By leveraging deep learning, the evaluation model is enhanced and constructed. Firstly, the system is refined by integrating the concept of integrating industry and education. Secondly, the traditional back propagation neural network is improved, and the model is built by incorporating deep noise reduction auto-encoder and support vector regression technology. Model test showed that model used in this study had good iterative performance. When the number of iterations was 62, the model started to enter a stable state, and the optimal fitness value in the stable state was 0.25. In addition, the detection accuracy was up to 0.98, and prediction effect can be satisfied by most teachers and students. To sum up, the quality evaluation model can accurately evaluate teaching quality, provide a reliable reference for colleges and enterprises, and promote the in-depth integration development of industry and education.