A Study of Campus Card Spending Big Data for Improving the Accuracy of Student Financial Aid

In the era of big data, colleges and universities can mine information from massive campus payment data to better implement accurate financial aid. In this paper, a machine learning model is constructed based on the campus card consumption data, a campus consumption database is established, hypothesis testing is used to screen the predictor variables, and Bagging, Boosting and Random Forest methods are applied to establish an integrated model based on decision trees to predict the economic status of students' families. It is found that the random forest model has the highest accuracy and is more effective in identifying economically disadvantaged students. This paper adopts computer modeling technology and proposes a new model of using consumption data to mine poor students, which provides a new path for exploring the accurate financial aid in colleges and universities.