Abstract
Higher education plays an important role in the improvement of people's quality and the development of our country. Therefore, it is necessary to evaluate the higher education curriculum. This paper analyzes and constructs the deep network learning system and self-encoder and evaluates the Chongqing higher education curriculum based on the deep learning network selected by 50 universities in Chongqing. It is found that the numbers of test objects, indicators, and hidden layers have an impact on the evaluation results. At the same time, a classroom teaching model is designed to improve the quality of higher education and solve the problem of insufficient curriculum quality of higher education.