Figure 3From: Prediction intervals for future BMI values of individual children - a non-parametric approach by quantile boosting Results from the linear setup, high-dimensional scenario, quantile boosting: Boxplots display the empirical distribution of the estimated coefficients for q 0:025 (left) and q 0:975 (right), obtained from 100 simulation runs. The blue lines represent the underlying true coefficients without shrinkage.Back to article page