Ensembles of Probabilistic Regression Trees
Abstract
Tree-based ensemble methods such as random forests, gradient-boosted trees, and Bayesian additive regression trees have been successfully used for regression problems in many applications and research studies. In this paper, we study ensemble versions of probabilistic regression trees that provide smooth approximations of the objective function by assigning each observation to each region with respect to a probability distribution. We prove that the ensemble versions of probabilistic regression trees considered are consistent, and experimentally study their bias-variance trade-off and compare them with the state-of-the-art in terms of performance prediction.
Domains
Machine Learning [stat.ML]Origin | Files produced by the author(s) |
---|