000002324 001__ 2324
000002324 005__ 20250829130901.0
000002324 0247_ $$2doi$$a10.6082/uchicago.2324
000002324 041__ $$aeng
000002324 245__ $$aXBART: A Scalable Stochastic Algorithm for Supervised Machine Learning with Additive Tree Ensembles
000002324 260__ $$bUniversity of Chicago
000002324 269__ $$a2020-06
000002324 300__ $$a107
000002324 336__ $$aDissertation
000002324 502__ $$bPh.D.
000002324 520__ $$aThis dissertation develops a novel stochastic tree ensemble method for nonlinear regression, which I refer to as XBART, short for Accelerated Bayesian Additive Regression Trees. By combining regularization and stochastic search strategies from Bayesian modeling with computationally efficient techniques from recursive partitioning approaches, the new method attains state-of-the-art performance: in many settings it is both faster and more accurate than the widely-used XGBoost algorithm. Via careful simulation studies, I demonstrate that our new approach provides accurate point-wise estimates of the mean function and does so faster than popular alternatives, such as BART, XGBoost and neural networks (using Keras). This dissertation also prove a number of basic theoretical results about the new algorithm, including consistency of the single tree version of the model and stationarity of the Markov chain produced by the ensemble version. Furthermore, I demonstrate that initializing standard Bayesian additive regression trees Markov chain Monte Carlo (MCMC) at XBART-fitted trees considerably improves credible interval coverage and reduces total run-time.
000002324 542__ $$fCC BY
000002324 650__ $$aStatistics
000002324 650__ $$aComputer science
000002324 653__ $$aBayesian
000002324 653__ $$aMachine Learning
000002324 653__ $$aMarkov chain Monte Carlo
000002324 653__ $$aRegression Trees
000002324 653__ $$aSupervised Learning
000002324 653__ $$aTree ensembles
000002324 690__ $$aBooth School of Business
000002324 691__ $$aBooth School of Business Dissertations
000002324 7001_ $$aHe, Jingyu$$uUniversity of Chicago
000002324 72012 $$aP. Richard Hahn
000002324 72012 $$aNicholas G. Polson
000002324 72014 $$aTengyuan Liang
000002324 72014 $$aRuey S. Tsay
000002324 8564_ $$95b6f8b63-a594-4d04-8cbe-8c81ea26d8b9$$s546795$$uhttps://knowledge.uchicago.edu/record/2324/files/He_uchicago_0330D_15292.pdf$$ePublic
000002324 909CO $$ooai:uchicago.tind.io:2324$$pDissertations$$pGLOBAL_SET
000002324 983__ $$aDissertation