Research on Predictive Model Technology for Student Academic Development Based on Machine Learning

Student performance management is a pretty significant part of campus service construction in colleges, there are many students who fail or even delay their graduation because of their low grades every year. If we can provide warning for students' grades and realize academic support early, we can reduce the failure rate and delayed graduation rate, improve the quality of campus services and enhance the level of school management. In this paper, a machine learning based student performance prediction model is established. Through mathematical analysis, seven kinds of easily accessible prediction parameters for schools are selected after comprehensive examination, and two classical machine learning prediction algorithms, KNN and random forest, are built to solve the problem. After substitution into the data set for training, better prediction results were achieved. As the amount of student data and data dimensions received by the school increase, the accuracy and generalization ability of the model will be continuously improved to finally realize the prediction model of student performance based on big data.