Evaluating regression model performance

In this section, we will review the regression evaluation measures for understanding the performance level of a regression model. Model evaluation is one of the key aspects of any machine learning model building process. It helps us to assess how our model will perform when we put it into production. We will use the following metrics for model evaluation:

• R-squared
• MSE
• MAE
• RMSE

R-squared

R-squared (or coefficient of determination) is a statistical model evaluation measure that assesses the goodness of a regression model. It helps data analysts to explain model performance compared to the base model. Its value lies between 0 and 1. A value near 0 represents a poor model while a value near 1 represents a perfect fit. Sometimes, R-squared results in a negative value. This means your model is worse than the average base model. We can explain R-squared using the following formula:

Let's understand all the components one by one:

• Sum of Squares...