Research on predicting student status changes based on machine learning methods:taking a local undergraduate university as an example
The article takes the academic performance of students in the 2018,2019,2020,and 2021 grades of local undergraduate colleges and universities as the research object.With different machine learning methods,the hidden data information is analyzed from two perspectives:the secondary college and the overall situation of the entire school.The results show that there is a significant difference in accuracy rate among different models.Through comparison,the LightGBM model has a relatively high accuracy rate of over 98%,and an AUC value of 0.811 6,indicating that the LightGBM model has the best predictive performance.Therefore,the lightGBM model is applied to the prediction of student status changes in universities for the 2021 grade.The prediction results indicate that the model can effectively predict student status changes,provide auxiliary decision-making advice for school academic management,improve teaching quality,and further promote the development and stability of the school.
local undergraduate colleges and universitiesmachine learningstudent status change
NETL
NSTL
万方数据
