Educational Big Data Mining and Analysis Based on Support Vector Machine Performance Prediction Rate

In this era, educational data has generated data-driven insights, especially with the application of machine learning models to predict the student performance. Even though previous researchers suggested various robust machines learning techniques still there are several challenges such as data complexity, privacy concerns, and the varied formats of educational data for efficient data processing and analysis. This study focuses on developing a robust model for predicting student performance by employing advanced machine learning methods tailored for educational data. This paper proposes a novel approach namely GIWRF-SVM utilizing Gini Impurity based Weighted Random Forest (GIWRF) with Support Vector Machine (SVM) performance prediction rate. Initially, the data is collected from Academic Performance Dataset (xAPI-Edu-Data) i.e. XAPI dataset. Then, the collected data is processed using various techniques such as discretization and resampling. Further, the processed data is fed as input to Gini Impurity based Weighted Random Forest (GIWRF) for feature selection. In the end, SVM is used for prediction of academic warning for college student's has acquired greater results with Mean Absolute Error (MAE) of 9.24, Root Mean Square Error (RMSE) of 4.28 and Root Absolute Error (RAE) of 4.76 when compared with Genetic Algorithm with Decision Tree Regression.