A number characterising the model prediction quality ( goodness of fit ):
R^2 = 1 - \frac{MSE(x, \hat x)}{MSE(x, \bar x)}, \quad 0 \leq R^2 \leq 1 |
where
a variable represented by a discrete data set of numerical samples | |
predictor of variable with the same number of samples | |
mean value of the variable | |
mean square error between a variable | |
mean square error between a variable |
The coefficient of determination normally ranges between 0, indicating a poor fit and 1, indicating a good fit.