高校学生课程绩点预测方法

On the Predictive Method of Course Grade Point for College Students

学业预警是加强学风建设、提高教学质量的一项举措,课程绩点预测是其中一项必不可少的工作.在分析课程之间的关系基础上,运用遗传优化BP神经网络算法对目标课程绩点进行预测.首先,将百分制成绩转化为课程绩点;其次,引入专业知识网络并建立课程社团;然后,运用关联规则分析给出课程之间的关联性;接下来,利用遗传算法优化BP神经网络,提出预测课程绩点的算法;最后,以本校数学专业成绩为分析数据,讨论该预测方法的可行性和有效性.

Academic performance pre-warning is an important method to strengthen the construction of academic style and improve the quality of teaching where the course grade prediction is absolutely necessary.Based on the analysis of the inter-curricular relationships,the optimized BP neural network genetic algorithm is used to predict the course grade point.Firstly,through introducing the course grade point,the percentage system is converted into the course grade point.Secondly,a professional knowledge network is introduced and a course community is created.Then,those strong association between courses are presented by aid of the association rules.Fourthly,an algorithm to predict the course grade by means of genetic algorithm to optimize the BP neural network is put forth.Finally,the score data of mathematics major in our university is analyzed to illustrate the feasibility and effectiveness of our method.