[Statistics] R-squared

[ R-squared  ]

R-squared  - a measure of the goodness-of-fit of a regression model. 

- It represents the proportion of the variance in the dependent variable that is predictable from the independent variables. 

- the percentage of variation explained by the relationship between two variables. 

 

=> range : 0 to 1 

=> R² = 1 - (SSR/SST)

SSR = the sum of squared residuals (the sum of the squared differences between the observed and predicted values of the dependent variable)

SST = the total sum of squares (the sum of the squared differences between the observed dependent variable and its mean)

 

-> High R-squared leads overfitting ! 

-> Overfitting: When a model is overfitted to the data, it can result in poor predictive performance and a negative R-squared

-> negative R-squared doesn't necessarily mean that the model is meaningless or useless. However, it does indicate that the model is not a good fit for the data and should be interpreted with caution