How ayushkumar418/student_performance_predictor works
This is a Streamlit student exam-score prediction product, with simple and advanced dashboard variants. The main workflow is student-entered factors → model prediction → score interpretation and personalized recommendations; the simple app also supports next-semester projections from student records. The advanced app adds confidence, feature-importance, dataset-analytics, and model-comparison views. Retraining and analysis are represented as separate supporting workflows. Some source excerpts are partial, so uncertain data paths and the listed prediction-history database are omitted.
Main entry points: app_advanced.py Advanced dashboard · train_advanced.py Training pipeline · model_analysis.py Dataset analysis · StudentPerformanceFactors.csv Performance dataset
Components 12
Prediction dashboards
- Simple dashboard
app.py - Advanced dashboard
app_advanced.py
Model lifecycle
- Selected model
- Model collection
- Training pipeline
train_advanced.py - Model metrics
model_results.json
Analytics and insights
- Feature rankings
feature_importance.json - Residual statistics
residuals.json - Dataset analysis
model_analysis.py - Analysis summary
analysis_summary.json
Student data
- Performance dataset
StudentPerformanceFactors.csv
Other components
- Student or educator
Connections 21
- Student or educator → Simple dashboard: enters factors
- Student or educator → Advanced dashboard: enters factors
- Simple dashboard → Selected model: loads and predicts
- Advanced dashboard → Selected model: loads and predicts
- Simple dashboard → Performance dataset: reads records
- Advanced dashboard → Performance dataset: reads for analytics
- Training pipeline → Performance dataset: loads and processes
- Training pipeline → Selected model: saves best model
- Training pipeline → Model collection: saves models
- Training pipeline → Model metrics: writes metrics
- Training pipeline → Feature rankings: writes rankings
- Training pipeline → Residual statistics: writes residuals
- Advanced dashboard → Model collection: loads for comparison
- Advanced dashboard → Model metrics: reads comparison metrics
- Advanced dashboard → Feature rankings: reads feature rankings
- Advanced dashboard → Residual statistics: reads confidence data
- Dataset analysis → Performance dataset: analyzes records
- Dataset analysis → Model collection: loads models
- Dataset analysis → Selected model: loads selected model
- Dataset analysis → Analysis summary: writes findings
- Advanced dashboard → Analysis summary: reads analysis