Root mean squared error (RMSE) measures prediction error after squaring each residual, averaging the squared errors, and taking the square root. Squaring makes large misses count disproportionately, so RMSE exposes painful failures that an average-error metric can hide. section 4.5.4
Use RMSE alongside MAE rather than treating it as a complete accuracy verdict. A much larger RMSE than MAE signals occasional extreme errors, while a lower RMSE does not automatically mean the model is better for every operational goal. section 4.5.5 section 4.5.6
Drug-price prediction
A Random Forest baseline produced an MAE of 77.50 INR and an RMSE of 179.69 INR. The gap suggests that most predictions were tolerable but some drug prices were missed by hundreds or thousands of rupees. An improved model reduced MAE to 57.98 INR while RMSE slightly increased to 181.50 INR, making the tradeoff visible instead of hiding it behind one score. section 4.5.5 section 4.5.6
A disciplined RMSE review should:
- Compare RMSE with MAE to distinguish routine error from occasional severe failure. section 4.5.4
- Inspect predicted-versus-actual or residual plots to locate under-prediction and high-variance regions. section 4.5.4
- Test proposed changes in isolation before accepting a lower average error as an overall improvement. section 4.5.6
[2] Model Evaluation (Random Forest Regressor): R-squared (R²): 0.3744 (How much variance is explained by the model) Mean Absolute Error (MAE): 77.50 INR (Average error in price prediction) Root Mean Square Error (RMSE): 179.69 INR
Review the baseline evaluation listing to compare R², MAE, and RMSE in one place. section 4.5.5









