Which metric is commonly used for regression models?

Mean Squared Error (MSE) is a standard and widely used metric for evaluating the performance of regression models. It measures the average of the squared differences between the predicted values and the actual values. A lower MSE indicates a better fit of the model to the data, as it signifies that the model's predictions are, on average, closer to the true values. It is particularly sensitive to large errors due to the squaring term.

A. Accuracy measures the proportion of correct classifications and is used for classification models, not regression models.

B. Precision measures the fraction of relevant instances among the retrieved instances and is a classification metric.

D. Recall (or sensitivity) measures the fraction of relevant instances that were retrieved and is a classification metric.

1. Microsoft. (2023). Evaluate model performance in Azure Machine Learning. Microsoft Azure Documentation. Retrieved from https://learn.microsoft.com/en-us/azure/machine-learning/how-to-understand-automated-ml-results. (In the "Regression/forecasting metrics" section, Mean Squared Error is listed and defined as a key metric for regression tasks.)

2. James, G., Witten, D., Hastie, T., & Tibshirani, R. (2013). An Introduction to Statistical Learning. Springer. (Chapter 3, Section 3.1.3, "Assessing the Accuracy of the Model," introduces Mean Squared Error as a primary method for quantifying the extent to which a predicted response value is close to the true response value in regression.)