基于隐语义模型的学生选课推荐算法

为了使学生可以准确、合理的进行选修课程,并调动其学习主动性,考虑到学生-课程之间潜在关系,提出了一种基于Funk-SVD技术的隐语义模型学生选课推荐算法.本算法使用随机梯度下降法优化损失函数;对选课推荐算法执行过程中的冷启动问题提出了一种处理方案;通过评价指标召回率、准确率以及平衡F分数验证本算法推荐的可行性和有效性,在所收集到的学生选课数据集上进行测试,实验结果表明,该算法具有一定的优势.

Recommended Algorithm for Students Course-choosing Based on Latent Factor Model

In order to enable students to take courses correctly and reasonably, and to arouse their enthusiasm of learning, in view of the actual relationship between students and courses, this thesis proposes a latent factor model of recommended algorithm for students on the basis of Funk-SVD technology. This algorithm applies a method of stochastic gradient descent to optimize the loss function; a solution to solve the problem of cold boot during the process of recommended algorithm for students course-choosing is provide accordingly; the feasibility and validity of this kind of recommended algorithm are verified by evaluating the index recall rate, accuracy rate, and balanced F score, testing on the data collected from students'' course-choosing. The experimental results show that the algorithm is advantageous.