Table 1.

Mean (SD) of $R_{\mathit{BS}}^2\left(2\right)$ for the training sets over 300 simulations

Selection approach	Weak signal		Strong signal
	n = 150	n = 300	n = 150	n = 300
ISIS-LASSO	0.122 (0.109)	0.578 (0.071)	0.335 (0.318)	0.797 (0.071)
ISIS-ALASSO	0.108 (0.108)	0.578 (0.071)	0.337 (0.322)	0.797 (0.071)
ISIS-RSF	0.069 (0.088)	0.564 (0.105)	0.232 (0.284)	0.790 (0.086)
SIS	0.065 (0.085)	0.303 (0.116)	0.114 (0.119)	0.491 (0.177)
SIS-LASSO	0.112 (0.093)	0.306 (0.113)	0.142 (0.119)	0.491 (0.177)
SIS-ALASSO	0.093 (0.092)	0.306 (0.113)	0.137 (0.120)	0.491 (0.177)
SIS-RSF	0.063 (0.082)	0.294 (0.125)	0.097 (0.104)	0.482 (0.178)
PSIS	0.756 (0.116)	0.811 (0.038)	0.642 (0.183)	0.869 (0.047)
PSIS-LASSO	0.875 (0.051)	0.805 (0.037)	0.869 (0.062)	0.874 (0.049)
PSIS-ALASSO	0.874 (0.052)	0.780 (0.042)	0.869 (0.063)	0.849 (0.057)
PSIS-RSF	0.688 (0.088)	0.413 (0.109)	0.655 (0.111)	0.555 (0.167)
LASSO	0.758 (0.138)	0.834 (0.043)	0.787 (0.136)	− 0.211 (0.036)
ALASSO	0.583 (0.115)	0.611 (0.058)	0.812 (0.095)	0.815 (0.063)

SD standard deviation