Statistics - Deviance


The deviance is negative two times the maximized log-likelihood.

And in the case of least squares regression, the deviance and the residual sum of squares are equivalent but for other model types the deviance is really just a generalization of residual sum of squares.

Powered by ComboStrap