Having covered metrics for classification tasks, this chapter focuses on evaluating models designed for regression problems. Regression models predict continuous numerical outcomes, such as predicting the price of a product or forecasting sales figures.

In this chapter, you will learn about common metrics used to quantify the performance of regression models. We will examine:

Mean Absolute Error (MAE): Measures the average magnitude of the errors in a set of predictions, without considering their direction.
Mean Squared Error (MSE): Measures the average of the squares of the errors. It gives higher weight to larger errors.
Root Mean Squared Error (RMSE): The square root of the MSE, providing an error metric in the same units as the target variable.
Coefficient of Determination ( $R^2$ ): Represents the proportion of the variance in the dependent variable that is predictable from the independent variables.

Understanding these metrics is fundamental for assessing how close a regression model's predictions are to the actual values. We will cover the calculation and interpretation of each metric to help you gauge model effectiveness for regression tasks.

Sections

3.1 Understanding Regression Predictions
3.2 Calculating Prediction Errors
3.3 Mean Absolute Error (MAE)
3.4 Mean Squared Error (MSE)
3.5 Root Mean Squared Error (RMSE)
3.6 Comparing MAE, MSE, and RMSE
3.7 Coefficient of Determination (R-squared)
3.8 Interpreting R-squared Values
3.9 Limitations of R-squared
3.10 Practice: Calculating Regression Metrics