<Dl> <Dd> F (θ) = ∫ d r g (r) ln ⁡ f (r; θ) (\ displaystyle F (\ theta) = \ int (\ textrm (d)) r \, g (r) \, \ ln f (r; \ theta)) </Dd> </Dl> <Dd> F (θ) = ∫ d r g (r) ln ⁡ f (r; θ) (\ displaystyle F (\ theta) = \ int (\ textrm (d)) r \, g (r) \, \ ln f (r; \ theta)) </Dd> <P> where g (r) (\ displaystyle g (r)) is the probability density of a random variable R (\ displaystyle R \,), and f (r; θ) (\ displaystyle f (r; \ theta) \,) with θ ∈ Θ i (\ displaystyle \ theta \ in \ Theta _ (i)) (i = 0, 1 (\ displaystyle i = 0, 1 \,)) are two families of parametric models . Model family 0 is the simpler one, with a restricted parameter space Θ 0 ⊂ Θ 1 (\ displaystyle \ Theta _ (0) \ subset \ Theta _ (1)). </P> <P> Parameters are determined by maximum likelihood estimation, </P>

When is the explained variation equal to 0