Table 2.

Summary of the modeling approaches included in the evaluation

Model	Ensemble characteristics		Output	Paradigm	R Package
	Tuning parameter	Model space construction
ENC	λ ENC	None	Influential variables	$\left.\right\}\text{Frequentist}(l_1,l_2\text{penalties})$	quadrupen, glmnet
PS	λ MB	$\left.\text{Subsampling}\right\}$	Influential variables
LS	λ ENC		Inclusion probabilities
SS	Λ		Inclusion probabilities
PR	λ MB	$\left.\right\}\text{Resampling}$	Influential variables
LR	λ ENC		Inclusion probabilities
SR	Λ		Inclusion probabilities
BMA	E M S=1	$\left.\right\}\text{MCMC}$	Inclusion probabilities	$\left.\right\}\text{Bayesian (Spike \& slab prior)}$	BoomSpikeSlab
BMAC	E M S CV		Inclusion probabilities

ENC: The baseline penalized regression model. Elastic net with λ _optimal=λ _ENC derived from cross-validation (CV), Ensembles based on 100 subsamples: PS: Meinshausen & Bühlmann’s algorithm with a single λ _optimal=λ _MB selected to minimize the expected number of false positives, LS: Single λ _optimal=λ _ENC with no variable selection, SS: Stability selection across the entire 100 λ∈Λ grid with no variable selection, Ensembles based on 100 resamples: PR, LR, SR: Identical to PS, PR and LR, respectively, with model space constructed through resampling. BMA: Bayesian model averaging with expected model size (EMS) = 1, BMAC: BMA with EMS determined by CV (E M S _CV).