Table 2.

Comparison of the proposed method and the Lasso in simulations

Setting	(p,n,σ)	Method	MSE	R 2	RE	FR
1	(20, 50, 0.5)	Lasso 1	0.29	0.89	0.77	0.33
			(0.01)	(0.00)	(0.01)	(0.00)
		Lasso 2	0.31	0.88	0.14	0.00
			(0.01)	(0.00)	(0.01)	(0.00)
		Proposed	0.29	0.89	0.13	0.00
			(0.01)	(0.00)	(0.01)	(0.00)
1	(100, 25, 0.2)	Lasso 1	0.19	0.92	0.80	0.37
			(0.07)	(0.03)	(0.02)	(0.02)
		Lasso 2	0.24	0.89	0.22	0.06
			(0.08)	(0.03)	(0.03)	(0.03)
		Proposed	0.19	0.92	0.21	0.04
			(0.07)	(0.03)	(0.03)	(0.03)
2	(20, 50, 0.5)	Lasso 1	0.32	0.89	0.44	0.20
			(0.01)	(0.00)	(0.02)	(0.04)
		Lasso 2	0.36	0.88	0.14	0.00
			(0.02)	(0.01)	(0.01)	(0.00)
		Proposed	0.32	0.89	0.14	0.01
			(0.01)	(0.00)	(0.01)	(0.01)
2	(100, 25, 0.2)	Lasso 1	0.42	0.84	0.52	0.37
			(0.07)	(0.02)	(0.03)	(0.06)
		Lasso 2	0.63	0.76	0.32	0.22
			(0.09)	(0.03)	(0.03)	(0.05)
		Proposed	0.41	0.85	0.31	0.24
			(0.07)	(0.02)	(0.03)	(0.05)
3	(20, 50, 0.5)	Lasso 1	0.30	0.93	0.18	0.00
			(0.01)	(0.00)	(0.02)	(0.00)
		Lasso 2	0.33	0.92	0.11	0.00
			(0.01)	(0.00)	(0.01)	(0.00)
		Proposed	0.30	0.93	0.11	0.00
			(0.01)	(0.00)	(0.01)	(0.00)
3	(100, 25, 0.2)	Lasso 1	0.15	0.96	0.25	0.00
			(0.02)	(0.01)	(0.03)	(0.00)
		Lasso 2	0.26	0.93	0.16	0.01
			(0.08)	(0.02)	(0.02)	(0.01)
		Proposed	0.15	0.96	0.15	0.00
			(0.02)	(0.01)	(0.01)	(0.00)

Sample means (standard errors in parentheses) of mean squared error (MSE), R ², relative error (RE) and false identification rate (FR), based on 20 simulation replications, for the proposed method and the Lasso