An ML engineer is building a model in SageMaker Canvas to make continuous numeric predictions using 10 years of historical data. Which evaluation metric should the engineer use to assess the model’s predictive performance?

Root Mean Square Error (RMSE) is the appropriate metric because the problem involves continuous numeric predictions, which is a regression task. RMSE measures the average magnitude of the errors between predicted and actual values, penalizing larger errors more significantly. Accuracy is incorrect because it's used for classification problems, where the output is a discrete class, not a continuous number. InferenceLatency measures the time it takes for a model to make a prediction, which is a performance metric, not an evaluation metric for predictive quality. Area Under the ROC Curve (AUC) is also a classification metric, used to evaluate the performance of binary classifiers across various threshold settings.