Google Cloud Professional Machine Learning Engineer Practice Test - Prep & Practice Exam

The correct answer highlights that MAE (Mean Absolute Error), MAPE (Mean Absolute Percentage Error), RMSE (Root Mean Square Error), RMSLE (Root Mean Square Logarithmic Error), and R² (coefficient of determination) are specifically designed to evaluate the performance of regression models. These metrics are crucial for assessing how well a model predicts continuous numerical outputs.

MAE provides a straightforward average of absolute errors, while MAPE gives insights into the percentage error relative to actual values, making it particularly useful for understanding performance across different scales. RMSE emphasizes larger errors more heavily due to squaring the loss, which can be important in specific applications where large deviations are particularly undesirable.

R² helps to understand how well the variation in the output can be explained by the model, indicating its predictive power. All of these metrics are essential for a linear regression context, where the goal is to estimate a target variable that is continuous.

The other types of metrics mentioned in the alternatives do not suit this context because classification metrics are intended for evaluating models that predict categorical outcomes rather than continuous values. Clustering metrics assess the quality of cluster assignments, while time series metrics focus on temporal data forecasting. Therefore, these alternatives do not align with the characteristics and application of