Research on the teaching quality evaluation of neural network machine learning algorithm

Abstract

Nowadays, there is no scientific and reasonable evaluation method for teachers teaching evaluation in higher education in China. According to the accumulated teaching experience, in order to ensure the fairness and perfection of teacher teaching evaluation, mathematical methods will be introduced into the evaluation work, such as analytic hierarchy process, grey decision method, fuzzy evaluation method, traditional statistical analysis evaluation model. Because teacher teaching evaluation is a nonlinear problem, the mathematical method has limitations in the application period, and both the selection index and the weight value are subjective. Therefore, on the basis of understanding the neural network algorithm, this paper constructs the teacher teaching evaluation system to think about problems from the perspective of different disciplines and specialties. The final experimental results show that using BP neural network for training and testing can further improve the rationality and objectivity of teachers teaching evaluation model.