In this paper, through the study of the policy and frontier theoretical changes in the educational evaluation of students and teachers in colleges and universities in recent years, a comprehensive analysis is made to summarize the primary indicators about educational evaluation, and then subdivide the secondary indicators downwards to quantitatively analyze and evaluate their rationalization. In order to improve the accuracy and scientific of education evaluation, this study proposes an AHP-BP education evaluation rationalization model based on the fusion of hierarchical analysis (AHP) and BP neural network. Firstly, the hierarchical analysis method is used to construct an evaluation index system, construct a comparison matrix, and after the initial weights of the evaluation indexes are derived from a two-by-two comparison, they are used as the input of the BP neural network. Finally, the BP neural network simulation results are used, and the model quantifies the concept of teaching evaluation indexes into definite data as the input of the network, and the simulation comprehensive evaluation results are used as the output. The hierarchical analysis method has strong subjective factors, and the neural network, by simulating the working principle of the human brain, has functions such as autonomous learning and non-linear transformation, which can effectively exclude randomness and subjectivity, and is an innovation in conducting educational evaluation. Therefore, the learning simulation process of BP neural network eliminates most of the subjective factors of the hierarchical analysis method in the process of continuous iteration, and obtains satisfactory evaluation results It has wide applicability. Finally, we conducted a small sample simulation based on some of the data collected on the Internet, and the results obtained were relatively small in error and met expectations, and the educational evaluation model could be put into use. Based on the AHP-BP education evaluation rationalization model and the understanding of the current situation of education evaluation, three main suggestions are made: 1. Encourage university evaluation to pay more attention to the process, and eliminate utilitarian evaluation indicators should be more decentralized indicator scores; 2. Suggest that the state should supervise in the general direction, and the specific situation of specific schools is different, and make moderate adjustments to the evaluation model; 3. Vigorously support teacher and student co-education The policy of activities to create a corresponding educational atmosphere resonates with the community and can promote the healthy development of university education.

Meili Wang. AHP-BP fusion model based on university education evaluation model and proposal. International Journal of New Developments in Education (2023) Vol. 5, Issue 3: 78-82. https://doi.org/10.25236/IJNDE.2023.050315.

[1] Calache L, Belotti C, Junior F, et al. Supplier selection and evaluation model proposal for industrial maintenance based on Fuzzy-TOPSIS [J]. Gestão & Produção, 2019, 26(2).

[2] Jiang Hua. Evaluation of university teachers from the perspective of high quality education system [J]. Journal of Hebei Normal University (Education Science Edition), 2022, 24(02):22-27. doi: 10.13763/j.cnki.jhebnu.ese. 2022.02.004.

[3] Guidance from the Ministry of Education on deepening the reform of the assessment and evaluation system of university teachers [EB/OL]. (2016-09-21) [2016-09-21]. http://www.gov.cn/xinwen/2016-09/21/content_5110529.htm

[4] Zhou Zheng. A review of the development status of BP neural networks [J]. Shanxi Electronic Technology, 2008(2): 90-92. 

[5] NIU Pengchao, SUN Yuan, SONG Wei, et al. Evaluation method of teachers' teaching ability based on BP neural network[C]// Artificial Intelligence and Security. ny, USA: ICAIS, 2019, 11632: 27-38. 

[6] Xiao, Y. & Guo, Chuan Yin, Application of hierarchical analysis in determining the weights of performance assessment indicators in refining and chemical design enterprises. Contemporary Chemical Industry, 2018. 47(02): pp. 396-399

[7] Sun Jiashuai et al. BP neural network based pressure drop and heat exchange prediction for oil coolers. Thermal Science and Technology, 2022: pp. 1-8.

[8] Yu, Feiyuan et al. Study on the risk prediction system of oil and gas pipeline ring weld failure based on BP neural network. Oil and Gas Field Surface Engineering, 2022. 41(04): pp. 71-77.

[9] Yu Min, Zhong Lei, Cao Hongbin. Research on practical teaching evaluation of universities based on AHP-BP neural network [J]. Foreign Economic and Trade,2022,(08):119-122.