Model report

Four regression algorithms were trained and compared per category group on a held-out 20% test split. Errors are in rupees on the real price scale (models are fit on log price). The best model by test RMSE is exported and used for live predictions.

electronics

Gradient Boosting selected

median error 12.82%

2,720 training rows · 681 test rows

Model	Test MAE	Test RMSE	Test R²	MAPE
Linear Regression	₹4,040	₹6,974	0.9316	14.8%
Random Forest	₹4,835	₹8,041	0.909	18.99%
Gradient Boosting	₹4,028	₹6,814	0.9347	14.75%
XGBoost	₹4,031	₹6,769	0.9355	14.55%

fashion

Linear Regression selected

median error 11.75%

2,116 training rows · 530 test rows

Model	Test MAE	Test RMSE	Test R²	MAPE
Linear Regression	₹2,300	₹4,489	0.9592	14.63%
Random Forest	₹3,175	₹6,229	0.9215	20.45%
Gradient Boosting	₹2,366	₹4,682	0.9557	14.8%
XGBoost	₹2,348	₹4,683	0.9556	14.99%

furniture

Linear Regression selected

median error 12.14%

504 training rows · 126 test rows

Model	Test MAE	Test RMSE	Test R²	MAPE
Linear Regression	₹2,970	₹5,293	0.9255	14.32%
Random Forest	₹4,037	₹6,824	0.8761	19.95%
Gradient Boosting	₹2,995	₹5,368	0.9234	14.58%
XGBoost	₹3,252	₹5,773	0.9113	15.77%

general

Gradient Boosting selected

median error 12.15%

2,161 training rows · 541 test rows

Model	Test MAE	Test RMSE	Test R²	MAPE
Linear Regression	₹1,129	₹2,223	0.9558	13.88%
Random Forest	₹1,469	₹2,793	0.9303	19.16%
Gradient Boosting	₹1,131	₹2,205	0.9565	14.36%
XGBoost	₹1,098	₹2,197	0.9568	13.9%

vehicles

Linear Regression selected

median error 13.25%

1,104 training rows · 276 test rows

Model	Test MAE	Test RMSE	Test R²	MAPE
Linear Regression	₹7,915	₹13,127	0.9613	14.55%
Random Forest	₹12,224	₹20,871	0.9022	22.41%
Gradient Boosting	₹8,498	₹14,169	0.9549	15.33%
XGBoost	₹8,654	₹14,909	0.9501	15.71%