EXPLAINABLE AI FOR STUDENT PERFORMANCE ANALYSIS IN CODING PLATFORMS

Abstract

Explainable AI (XAI) simplifies and clarifies complex machine learning models. Grading students' code platform performance becomes increasingly personal. Many earlier machines are "black boxes," making assumptions without explanation. This keeps teachers and students in the dark. XAI makes these predictions more transparent, so everyone can understand how diverse factors affect the final output. The system analyzes students' coding habits, submission trends, error types, and problem-solving tactics to discover what motivates and hampers them. SHAP and LIME simplify difficult model selections by emphasizing student work's most important parts. This shows your strengths and weaknesses, enabling more targeted feedback and help. These findings help educators identify at-risk students and improve class preparation and delivery. Finally, XAI supports fair, accountable, and practical educational analytics. This makes learning on coding platforms more welcoming and supportive.