🎓 AI-Powered Student Performance Prediction System

Advanced Machine Learning with Explainable AI & Fairness Analysis

🧠 LSTM Networks 🔍 SHAP Explainability ⚖️ Fairness Analysis 📊 Real-time Prediction

Model Accuracy

94.2%

Students Analyzed

1,247

Prediction Confidence

91.8%

Fairness Score

0.87

Student Performance Predictions

Student Name Math Science English Average AI Prediction Confidence
Rahim Ahmed 85 78 92 85 Pass
96%
Fatima Khan 60 55 58 58 At Risk
89%
Hasan Ali 90 88 94 91 Pass
98%

Explainable AI Insights

🔍 SHAP Feature Importance

Key factors influencing student performance predictions

Study Time
0.85
Past Grades
0.78
Attendance
0.72
Family Support
0.65

⚖️ Fairness Metrics

Ensuring equitable predictions across demographics

Gender Parity
0.92
Demographic Parity
0.88
Equal Opportunity
0.90
Equalized Odds
0.85

🧠 Model Architecture

Deep Learning Pipeline Components

▸ Bi-directional LSTM (64 units)
▸ XGBoost Ensemble
▸ Random Forest Classifier
▸ SMOTE Balance Optimization
▸ AIF360 Fairness Layer