Construction and Validation of a Neural Network-Based Blended Teaching Quality Evaluation Model

DOI: https://doi.org/10.62517/jbdc.202601211

Author(s)

Dongxue Tu1,2

Affiliation(s)

1Department of Basic Course Teaching and Research, Guangdong Technology College, Zhaoqing, Guangdong, China 2Faculty Of Arts, Communication & Education, Centre for Postgraduate Studies (CPS), Kuala Lumpur University of Science and Technology, Selangor, Kuala Lumpur, Malaysia

Abstract

Addressing the persistent challenges in blended teaching quality assessment—overly intricate indicator systems, pronounced subjectivity, and the inadequacy of conventional evaluation approaches—this study takes mathematics courses as a case study and proposes a neural network-based evaluation framework. From offline questionnaires and online teaching platforms, 118 valid samples were obtained. Following min-max normalization, two distinct models were constructed: a standard BP network configured with a 20-7-1 architecture, and a genetic algorithm-enhanced variant. The latter utilized the evolutionary algorithm to optimize initial connection weights and thresholds, resulting in a 20-41-1 topology. MATLAB was employed for both training and testing. The experimental findings show that the standard BP model produced an average error of 4.61 and a relative error of 0.05. By contrast, the model refined through genetic optimization reduced these figures to 1.94 and 0.02, respectively, while also demonstrating a markedly narrower error spread. Its predictive robustness surpassed not only that of the basic BP model but also that of the benchmark algorithms (GA and BSA). The outcomes confirm that the hybrid GA-BP approach achieves significantly higher accuracy and reliability for blended teaching quality evaluation, offering a dependable scientific basis for instructional management decisions.

Keywords

Blended Teaching; Teaching Quality Evaluation; Genetic Algorithm Optimization; Mathematics Courses

References

[1] Mei L. Model Construction of higher education quality assurance system based on fuzzy neural network. Informatica, 2024, 48(10). [2] Nan Z. Research on the application of BP algorithm of neural network in teaching quality evaluation //2024 International Conference on Optimization Computing and Wireless Communication (ICOCWC). IEEE, 2024: 1-4. [3] Huang Lu, Ren Hang, Liu Mingxi, et al. Research on the Measurement Method of Industry-University-Research Knowledge Flow Integrating Large Language Models and Dynamic Graph Neural Networks. Chinese Journal of Management Science, 2025, 1-16. [4] Li Ping, Wang Zhi, Pan Haijing. Construction of a Diversified Teaching Evaluation System for University Mathematics Courses Based on Learning Community. Forest Region Teaching, 2025, (10):97-100. [5] Zhang Hongyi, Tang Siwei. The Structure and Cultivation of Rural Teachers' Place-based Teaching Evaluation Competence. Journal of Teaching and Management, 2026, 1-5. [6] Yang H, Fu Y F. Construction of blended teaching quality evaluation indicator system for vocational colleges in the new era. Vocational and Technical Education, 2021, 42(35): 67-72. [7] Duan Wenwen. New Quality Productive Forces Empowering High-Quality Development of University Education and Teaching Management: A Review of 'Research on the Current Situation and Innovative Development of University Education and Teaching Management'. Higher Education Exploration, 2026, (03):2. [8] Xu Anfeng, Li Bin, Li Yue. Research on Online and Offline Blended Smart Education Model. Theory and Practice of Education, 2024, 44(15):57-60. [9] Jiang Hongling, Chen Linan. Construction and Practice of Online and Offline Blended Teaching Evaluation System: Taking Computer Network Course Teaching as an Example. China Modern Educational Equipment, 2026, (07):140-143+147. [10] Zhang Jingjing, Huang Yicheng. The Policy Implementation Logic of Artificial Intelligence Empowering Blended Teaching Reform. Ethnic Education Research, 2025, 36(06):31-41. [11] Lai Jun, Zhang Jinrui. Simple Implementation of Adaptive Fast Multipole Algorithm Based on MATLAB. Mathematica Numerica Sinica, 2025, 47(01):1-20. [12] Alhijawi B, Awajan A. Genetic algorithms: Theory, genetic operators, solutions, and applications. Evolutionary Intelligence, 2024, 17(3): 1245-1256.