Skip to main content
. Author manuscript; available in PMC: 2017 Dec 1.
Published in final edited form as: Psychol Methods. 2016 Dec;21(4):583–602. doi: 10.1037/met0000087

Figure 7.

Figure 7

Prediction performance of multivariate tree boosting, MANOVA, the Lasso, multivariate CART, and bagged multivariate CART. Lower prediction error is better. The performance is shown for a given nonlinear effect (ex, x2, x3, x is the linear model) for a range of effect sizes. Multivariate tree boosting has lower prediction error than bagged multivariate CART, MANOVA or the Lasso for conditions with nonlinear effects (x2, x3). It has comparable performance when effects are linear (x) or approximated well by a linear model (ex).