An Effective Method to Improve the Accuracy of Precise Funding in Campus

Abstract

In the existing precise subsidy system, how to accurately select the students who need to be subsidized by analyzing a large number of student campus consumption data has become a difficult problem. We analyzed the campus consumption data of 5505 students in detail, and visualized the consumption differences of students in different colleges, grades and genders. We proposed a combination of K-means clustering and conditional expression, taking full account of days at school, frequency of meals and average daily consumption. According to the analysis results of this algorithm, more than 100 students were funded. The experimental data and the feedback received indicate that the algorithm has better accuracy.