Fig. 4From: Randomized boosting with multivariable base-learners for high-dimensional variable selection and predictionResults for low-dimensional simulation setting. Boxplots of false positives, false negatives, estimation error and prediction error on test set (of size 1000), for 500 simulation replicates with \(n=100\), \(p=20\), \(S_{{{\text {true}}}}=\{1,2,3,4\}\) and Toeplitz correlation with \(\rho =0.8\)Back to article page