Construction and application of a performance assessment model for teaching operations research based on Bayesian network technology

Operations research plays a pivotal role in fostering effective management, sound decision-making, and modern management practices. Its significance in the academic realm, particularly within higher education, is increasingly acknowledged. This study leverages a Bayesian network to introduce a Markov chain, thereby constructing a dynamic Bayesian network-based model for evaluating the teaching performance of operations research. Initially, a questionnaire survey method was utilized to make preliminary adjustments to the performance evaluation index system. Subsequently, various methods, including the principle of halves and the coefficient of variation, were applied for coarse and fine screening of indices, culminating in the development of a comprehensive index system for evaluating operations research teaching performance. The efficacy of the constructed model and index system was tested by assessing the performance of five operations research instructors. Key metrics such as teaching performance, student trust, and student interest were assigned the highest weights, with respective values of 0.153, 0.127, and 0.114. Notably, Teacher D achieved the highest total performance score of 9.051, surpassing the other four instructors. The assessment results obtained from the designed model are closely aligned with those derived from fuzzy clustering techniques, underscoring the model’s robustness and applicability. Thus, the performance assessment model and index system designed for operations research teaching work demonstrate substantial practical effectiveness.