Academic Performance Analysis of the Students with Feature Retraction Using Machine Learning Algorithms

This research investigates the application of machine learning algorithms to predict student academic performance based on historical data and current factors. Five machine learning algorithms were employed to analyze thirteen features related to student demographics, academic background, and other relevant factors. The study evaluates the performance of the algorithms using classification metrics, including the classification matrix and confusion matrix. The results demonstrate that after data preprocessing, four out of the five classifiers achieved 100% accuracy, while K-Nearest Neighbors (KNN) achieved 80% accuracy. Random Forest (RF) exhibited a significant improvement from 92% accuracy before preprocessing to 100% accuracy after preprocessing. Chi-square analysis reveals that age and gender features has no significant impact on academic performance. This finding suggests that other factors, such as study habits, motivation, and socioeconomic conditions, may play a more influential role in predicting student success. The research contributes to the understanding of factors influencing student academic performance and provides valuable insights for educators and policymakers. By leveraging machine learning techniques, this study demonstrates the potential for predicting student success and implementing targeted interventions to support student achievement.