MACHINE LEARNING BASED STUDENT ACHIEVEMENT PERFORMANCE PREDICTION WEB APPLICATION

Abstract

The use of multiple linear regression in our research is critical for determining the factors that have a greater impact on student performance index. Machine learning studies that employ multiple linear regression models to forecast student performance index aim to increase educational processes and individual student ability. These studies search to gain a deeper understanding of the factors that impact academic success by examining various variables that affect student performance. In the literature has demonstrated that such models achieve high levels of accuracy and can reliably predict student performance. In our study, we constructed and trained a multiple linear regression model. The dataset was divided into training and test sets, and the model was assessed using these datasets. Performance of the model was evaluated using various metrics such as MAE, MSE, R2, RMSE, and Accuracy(ACC). The results obtained indicated that the model performed exceptionally well, indicating its ability to make precise predictions. Especially, the coefficient of determination (R2) was 0.99, and the ACC value was 0.994, underscoring the model's exceptional ability to accurately explain the data. The focus of our research is to assess the precision and dependability of the findings derived from analyzing the impact of different independent factors on student achievement, utilizing a multiple linear regression model. Moreover, we have created a web interface using the Flask web module that enables the prediction of student performance based on inputting new variables.