基于贝叶斯网络的研究生入学奖学金评定

研究生入学奖学金评价体系的建立是我国高校研究生培养机制改革中的重要问题。为了公平、公正地对研究生入学奖学金进行评定，需要把握影响奖学金等级评定的相关因素，并分析这些因素之间的内在关系。文中以历史数据为依据采用K2算法构建评定奖学金等级的贝叶斯网络模型，并基于概率推理算法对奖学金的等级进行预测。研究结果表明，该方法是可行的，其准确率高达88%，为研究生入学奖学金的评定提供了科学依据。

Research on Scholarship Evaluation for Graduates Admission Based on the Bayesian Network

The establishment of the system of graduate scholarships evaluation is one of the most important problems in the graduate education reform. In order toassess graduate scholarships more fairly and equitably, it is needed to analyze the relevant factors of scholarship levels and their intrinsic relationships. Based on historical data, a Bayesian network model for evaluation of scholarship levels is constructed by using the K2 algorithm and the probabilistic inference algorithms for prediction. The research results show that the proposed method is feasible and its accuracy is as high as 88%. The study presents a scientific method for the evaluation of graduate admission scholarship, which provides a good guidance for the future scholarship evaluation.