Predicting Graduation and Dropout Rates : A Machine Learning Approach

Abstract

Student dropout is a complicated and detrimental issue in the educational process, with costs to both students and institutions on a social and financial level. The tool presented in the paper uses machine learning techniques to forecast first-year undergraduate student dropouts. In order to calculate the risk of dropping an academic course, the tool takes into account personal information, secondary school academic records, and first-year course credits. In this paper, a broad study on the applications of the algorithms of machine learning is presented for predicting educational enrollment status. The research models in this situation. The used dataset includes a wide range of financial, academic, and demographic variables, offering a rich source of data for modeling. Researchers and practitioners looking to use similar predictive models in educational contexts can benefit greatly from the methodology and insights presented here.Focuses on comparing the efficacy of Support Vector Machine (SVM) and Gaussian Naive Bayes (GNB). The results of this study show that the SVM Model demonstrates higher accuracy than the GNB Model. Accuracy of these models then performs an increase when techniques like SMOTE is applied. This study lays the groundwork for future developments in the field of predictive modeling for student enrollment and adds to the body of knowledge in that area.

Ähnliche Arbeiten