TY - GEN
T1 - Bayesian Ensemble Learning
AU - Chipman, Hugh A.
AU - George, Edward I.
AU - McCulloch, Robert E.
N1 - Funding Information:
The authors would like to thank three anonymous referees, whose comments improved an earlier draft, and Wei-Yin Loh who generously provided the datasets used in the experiment. This research was supported by the Natural Sciences and Engineering Research Council of Canada, the Canada Research Chairs program, the Acadia Centre for Mathematical Modelling and Computation, the University of Chicago Graduate School of Business, NSF grant DMS 0605102 and by NIH/NIAID award AI056983.
Publisher Copyright:
© NIPS 2006.All rights reserved
PY - 2006
Y1 - 2006
N2 - We develop a Bayesian “sum-of-trees” model, named BART, where each tree is constrained by a prior to be a weak learner. Fitting and inference are accomplished via an iterative backfitting MCMC algorithm. This model is motivated by ensemble methods in general, and boosting algorithms in particular. Like boosting, each weak learner (i.e., each weak tree) contributes a small amount to the overall model. However, our procedure is defined by a statistical model: a prior and a likelihood, while boosting is defined by an algorithm. This model-based approach enables a full and accurate assessment of uncertainty in model predictions, while remaining highly competitive in terms of predictive accuracy.
AB - We develop a Bayesian “sum-of-trees” model, named BART, where each tree is constrained by a prior to be a weak learner. Fitting and inference are accomplished via an iterative backfitting MCMC algorithm. This model is motivated by ensemble methods in general, and boosting algorithms in particular. Like boosting, each weak learner (i.e., each weak tree) contributes a small amount to the overall model. However, our procedure is defined by a statistical model: a prior and a likelihood, while boosting is defined by an algorithm. This model-based approach enables a full and accurate assessment of uncertainty in model predictions, while remaining highly competitive in terms of predictive accuracy.
UR - http://www.scopus.com/inward/record.url?scp=85158060143&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85158060143&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85158060143
T3 - NIPS 2006: Proceedings of the 19th International Conference on Neural Information Processing Systems
SP - 265
EP - 272
BT - NIPS 2006
A2 - Scholkopf, Bernhard
A2 - Platt, John C.
A2 - Hofmann, Thomas
PB - MIT Press Journals
T2 - 19th International Conference on Neural Information Processing Systems, NIPS 2006
Y2 - 4 December 2006 through 7 December 2006
ER -