A New Ridge-Type Estimator for the Gamma Regression Model

Table 1

Estimated mean squared error when p = 4.

	n	ρ	MLE	GRE-k	GLE-d	GKL ()	GKL ()
0.5	20	0.95	2.008	0.949	1.643	1.193	0.942
		0.99	8.195	2.761	7.156	4.083	2.018
		0.999	78.599	23.305	75.070	37.119	17.929
	50	0.95	1.265	0.643	1.025	0.763	0.601
		0.99	4.277	1.257	3.532	1.799	1.102
		0.999	38.172	8.044	35.320	13.298	7.051
	200	0.95	0.544	0.444	0.478	0.459	0.435
		0.99	0.923	0.467	0.682	0.551	0.463
		0.999	5.068	0.554	4.067	1.522	0.545
1	20	0.95	3.514	1.758	3.113	2.025	1.357
		0.99	15.677	6.753	14.558	8.226	4.568
		0.999	154.076	63.790	150.439	79.217	61.203
	50	0.95	2.671	1.528	2.406	1.655	1.155
		0.99	11.034	5.410	10.200	6.003	2.205
		0.999	105.109	48.863	102.240	54.610	26.562
	200	0.95	0.628	0.449	0.546	0.473	0.445
		0.99	1.392	0.504	1.050	0.683	0.463
		0.999	9.837	3.220	8.355	2.948	1.276