Research and application of grade prediction model based on decision tree algorithm

With the rapid development of online open courses, it has become a hot issue in educational research to use learning analysis technology to explore online learning characteristics and provide support for the improvement of teaching methods and academic warning. A grade prediction model based on decision tree algorithm according to students' previous learning behavior is designed in this paper. Firstly, ID3, C4.5 and CART classification algorithm are used to construct the initial grade prediction model for the pre-processed learning behavior data set. Secondly, in order to prevent over-fitting, pruning set is used to conduct pruning on the initial grade prediction model. Finally, the performance of three grade prediction models is evaluated by test data set. Experimental results indicate that the grade prediction model constructed by ID3 algorithm has better performance than that constructed by other two algorithms.