End-to-end machine learning solution for predicting food delivery times with 94.77% accuracy (R² = 0.9477).
99.51% of test predictions within ±5 minutes of actual delivery time | 11,399 predictions generated
| Metric | Value |
|---|---|
| Records | 56,556 (45,157 train + 11,399 test) |
| Features | 37 engineered features |
| Model | Random Forest Regressor |
| Accuracy | R² = 0.9477 (94.77%) |
| Error | MAE = 1.70 min |
| Prediction Accuracy | 99.51% within ±5 min |
| Queries | 23 SQL, 22 CSV exports |
| Scripts | 8 production-ready |
| Visualizations | 6 charts |
# Data processing
python python/01_explore_data.py
python python/02_data_cleaning.py
python python/03_load_to_sql.py
# SQL Analytics
python python/04_run_sql_analytics.py
# ML Pipeline
python python/05_eda_analysis.py
python python/06_ml_model_training.py
python python/07_predictions.py
python python/08_model_evaluation.py📁 python/ - 8 production scripts 📁 sql/ - 23 queries, 22 exports 📁 data/ - Raw, cleaned, SQL data 📁 ml_models/ - Trained models + predictions 📁 docs/ - Reports, visualizations, guides
Total Files: 100+ | Total Size: ~150 MB
Random Forest (Best Model)
- R² Score: 0.9477
- RMSE: 2.08 minutes
- MAE: 1.70 minutes
- Accuracy: 99.51% within ±5 min
Top Features
- delivery_speed (94.98%)
- delivery_distance_km (0.81%)
- Delivery_person_Age (0.59%)
- Weatherconditions (0.44%)
- Vehicle_condition (0.43%)
✅ Data processing pipeline (37 engineered features) ✅ SQLite database (56,556 records) ✅ 23 SQL queries with business analytics ✅ EDA with 4 visualizations ✅ 3 trained ML models ✅ 11,399 test predictions ✅ Model evaluation & diagnostics ✅ 6-page Tableau dashboard design ✅ Complete documentation
📖 PROJECT_SUMMARY.md - Complete project overview (all 14 days) 📖 TABLEAU_DASHBOARD_GUIDE.md - Dashboard blueprint + implementation 📖 DELIVERABLES.md - Complete file listing & specifications
🎯 Build Tableau Dashboard using:
- 22 SQL query CSV files (ready)
- ML predictions (ready)
- 6 pre-designed pages (see guide)
- Implementation steps provided
Status: ✅ Production-Ready Last Updated: December 14, 2025 Duration: 14 Days Complete