Design and optimization of the Teaching evaluation system based on artificial intelligence

Abstract

Abstract In today’s highly developed information society, the construction work of educational informatization has become the focus of the future development of major universities. This paper combines artificial neural networks and UML software modeling technologies to design a teaching evaluation system using artificial neural network technology. The system requirements are analyzed from three aspects: system feasibility, functional requirements, and neural network evaluation function. The proposed system performance requirements are based on these three aspects. The neural network approximation problem can be solved using the gradient descent method based on the BP neural network model. After a series of operations, such as data preprocessing, the Design of the teaching evaluation system is completed, realizing the joint role of teaching and neural network evaluation module, and at the same time, proposing a system optimization decision for the teaching quality aspect. The teaching evaluation system designed in this study passed the test as the average response time of the whole system was 94.366 seconds when the number of users was 100, and there was no phenomenon such as transaction unresponsiveness. After the use of the teaching system, the feedback from the students’ use shows that the number of computer science students who mastered more than 80% of the system concepts is 42, and the rest of the knowledge is concentrated in the number of people who mastered 40% to 80%.