Note. Available from PSG Version 2.3

Pseudo_R2 shows quality of linear regression in the sense of PSG error function minimization.

 

Mathematical Definition

Pseudo_R2 is calculated as follows:

,

where

 is one of PSG Error functions (see Error Group),

 

is the residual, which is the Loss Function, defined on Matrix of Scenarios

,

is a point, e.g. an optimal point  of the minimization problem .

 

is  the residual when only intercept is included, with Matrix of Scenarios:

,

is an optimal point  of the minimization problem .