Generalized ridge estimator outperforms least square in regression accuracy.
The generalized ridge estimator of regression coefficients is better than the least square estimator in the Gauss-Markov model, according to the mean square error matrix and Pitman Closeness criteria.