A Poor College Students Identified Model Based on Decision Tree

In this paper, we propose a Poor College Students identified model based on decision tree. The model improves the accuracy of the poor students' identification and reduces the workload. According to the results, we propose a way to supplement the students with high consumption frequency and low consumption money. This is a good way to supplement the incomplete and inaccurate identification of poor students in colleges. First, we process campus card consumption data through data cleaning. Then, we introduce a list of poor students from the academic office, and select different attributes in the consumption data to form multiple training samples. Finally, the J48 algorithm in the Weka platform is used to analyze the sample data, and the college poor student identification model is obtained. Through the comparison and practice with the list of accredited laboratories, the accuracy of our model can reach about 70%, and the way of meal replenishment was also well received by teachers and students.