Variable Selection and Parameter Estimation with the Atan Regularization Method

Table 1

Simulation results for linear regression models of Example 8.

Method	MRME	Number of zeros		Proportion of
		C	IC	Underfit	Correct-fit	Overfit
,
Lasso	0.6213	3.0000	1.2050	0.0000	0.3700	0.6700
Alasso	0.3074	3.0000	0.3500	0.0000	0.7300	0.2700
SCAD	0.2715	3.0000	1.0650	0.0000	0.4400	0.5600
MCP	0.3041	3.0000	0.5650	0.0000	0.5800	0.4200
Dantzig	0.4623	3.0000	0.6546	0.0000	0.5700	0.4300
Bayesian	0.3548	3.0000	0.5732	0.0000	0.6300	0.3700
Atan	0.2550	3.0000	0.1750	0.0000	0.8450	0.1550
,
Lasso	0.6027	3.0000	1.0700	0.0000	0.3550	0.6450
Alasso	0.2781	3.0000	0.1600	0.0000	0.8650	0.1350
SCAD	0.2900	3.0000	0.8550	0.0000	0.5250	0.4750
MCP	0.2752	3.0000	0.3650	0.0000	0.6850	0.3150
Dantzig	0.3863	3.0000	0.8576	0.0000	0.4920	0.5080
Bayesian	0.2563	3.0000	0.4754	0.0000	0.7150	0.2850
Atan	0.2508	3.0000	0.1000	0.0000	0.9050	0.0950