Construction and application of evaluation index system for university teachers based on adaptive mutation algorithm

Abstract

Teaching evaluation of college teachers is an indispensable part of professional certification. In order to realize the scientific and objective evaluation of teaching quality, this paper constructs 18 secondary indicator systems covering teacher quality, teaching content, teaching attitude, teaching methods and teaching effects based on the six basic principles of professional certification. In order to improve the prediction ability of the evaluation model, the adaptive mutation genetic algorithm (AGA) is introduced to optimize the back propagation (BP) neural network, aiming to predict the evaluation results of teachers’ teaching quality. In addition, the objective weights of the evaluation indicators are calculated by the entropy method to simulate the actual evaluation process, which provides a reliable reference standard for the AGA-BP model. The experimental results show that the prediction error range of the AGA-BP algorithm is 0.02–0.03, which has higher accuracy than the error range of the traditional GA-BP algorithm (0.05–0.08). The study provides a feasible solution for the scientific evaluation of the teaching quality of college teachers and provides data support for professional certification.