Research on the prediction of the number of scholarships won by college students based on the RF-LightGBM model

Abstract

With the increasing number of university graduates every year, college students are facing serious employment challenges. Good academic performance and a higher number of scholarships are conducive to enhancing students' competitiveness, however, they lack the ability to plan their academic life reasonably. In order to guide them to plan their study life reasonably and efficiently to promote better employment, this study proposes the RF-LightGBM model for predicting the number of scholarships awarded to students and analysing the important factors affecting the number of academic scholarships awarded. The experimental results show that the accuracy of the model reaches 83.1%, which is higher than other models, and can predict the number of scholarships awarded more accurately. The key features affecting the number of scholarships awarded are the grades of professional courses, and the scores assigned to the behavioural performance in school, so it is suggested that students should pay attention to the study of professional courses, and take an active part in professional competitions, volunteer activities, etc., so as to improve the comprehensive quality of the individual. This study provides a feasible reference for students on how to better obtain academic scholarships, which will help them plan their study life and enhance their competitiveness so that they can better cope with the challenges of employment.