SuperKart Sales Regression Model
Model
Selected model: Random Forest
The model is implemented as a scikit-learn Pipeline containing feature preprocessing and the selected regression estimator.
Cross-Validation Performance
CV RMSE: 285.89
Test Performance
MAE: 105.10
RMSE: 278.11
R2: 0.9322
Hyperparameters
{'model__max_depth': None, 'model__min_samples_leaf': 2, 'model__min_samples_split': 5, 'model__n_estimators': 200}
Numerical Features
- Product_Weight
- Product_Allocated_Area
- Product_MRP
- Store_Establishment_Year
Categorical Features
- Product_Sugar_Content
- Product_Type
- Store_Id
- Store_Size
- Store_Location_City_Type
- Store_Type
Target
Product_Store_Sales_Total
Dataset Note
The supplied dataset does not contain a date or time variable. Therefore, this model performs product-store sales prediction using regression rather than chronological time-series forecasting.