Modified Gradient Boosting in High-dimensional Nonlinear Regression

<a href="https://embed.stanford.edu/iframe/?url=https%3A%2F%2Fpurl.stanford.edu%2Fhk301nk9163" class="su-underline">Show Content</a>

Abstract/Contents

Abstract: Herein we develop a modification of Freund and Schapire's (1997) AdaBoost algorithm for classification trees and Friedman's (2001) gradient boosting generalization. Not only does this modification address long-standing unresolved problems concerning convergence and when to terminate the iterative algorithms, but it is also shown to have optimality properties via attainment of the convergence rates of oracle benchmarks that are computationally infeasible. Simulation studies are presented to illustrate the key ideas underlying modified gradient boosting and how it works.

Subject	steepest-descent minimization of loss functions
Subject	orthogonal matching pursuit
Subject	semi-population model
Subject	weak greedy algorithms
Genre	Text
Genre	Technical report

DOI	https://doi.org/10.25740/hk301nk9163
Location	https://purl.stanford.edu/hk301nk9163

Use and reproduction: User agrees that, where applicable, content will not be used to identify or to otherwise infringe the privacy or confidentiality rights of individuals. Content distributed via the Stanford Digital Repository may be subject to additional license and use restrictions applied by the depositor.
License: This work is licensed under a Creative Commons Attribution Non Commercial No Derivatives 4.0 International license (CC BY-NC-ND).

Preferred citation: Lai, T., Xia, T., and Yuan, H. (2021). Modified Gradient Boosting in High-dimensional Nonlinear Regression. Department of Statistics Technical Report, Stanford University. Available from the Stanford Digital Repository at https://purl.stanford.edu/hk301nk9163

Statistics Department Technical Reports

Loading usage metrics...