Exam AWS Certified Machine Learning Engineer - Associate MLA-C01 topic 1 question 90 discussion

C. Root mean square error (RMSE) ✅ E. Average weighted quantile loss (wQL) Here's why: RMSE is a standard regression metric that measures the average magnitude of the error between predicted and actual values. It penalizes larger errors more than smaller ones. Weighted Quantile Loss (wQL) is commonly used in probabilistic forecasting models (e.g., Amazon Forecast, DeepAR). It assesses the accuracy of quantile predictions.

🔑 Keyword: Time-series forecasting model evaluation metrics ✅ Correct Answers: C. Root mean square error (RMSE) & E. Average weighted quantile loss (wQL) Why? RMSE (Root Mean Square Error): Measures error magnitude in time-series predictions. wQL (Weighted Quantile Loss): Used in Amazon Forecast to evaluate quantile forecasts. Why Others Are Wrong? ❌ A. Recall is a classification metric, not for forecasting. ❌ B. LogLoss is used for classification models, not time-series. ❌ D. Inference latency is about model performance, not accuracy.