Boosted nonparametric hazards with time-dependent covariates

DOI10.1214/20-AOS2028MaRDI QIDQ2054480zbMATH OpenFDO

Authors Donald K. K. Lee, Ningyuan Chen, Hemant Ishwaran

Publication date 3 December 2021

Published in The Annals of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1701.07926, https://projecteuclid.org/journals/annals-of-statistics/volume-49/issue-4/Boosted-nonparametric-hazards-with-time-dependent-covariates/10.1214/20-AOS2028.full

zbMATH Keywords

functional datasurvival analysisgradient boostingregression treeslikelihood functionalstepsize shrinkage

Mathematics Subject Classification ID

Nonparametric estimation (62G05) Asymptotic properties of nonparametric inference (62G20) Estimation in survival analysis and censored data (62N02) Functional data analysis (62R10)

Abstract: Given functional data from a survival process with time-dependent covariates, we derive a smooth convex representation for its nonparametric log-likelihood functional and obtain its functional gradient. From this, we devise a generic gradient boosting procedure for estimating the hazard function nonparametrically. An illustrative implementation of the procedure using regression trees is described to show how to recover the unknown hazard. The generic estimator is consistent if the model is correctly specified; alternatively, an oracle inequality can be demonstrated for tree-based models. To avoid overfitting, boosting employs several regularization devices. One of them is step-size restriction, but the rationale for this is somewhat mysterious from the viewpoint of consistency. Our work brings some clarity to this issue by revealing that step-size restriction is a mechanism for preventing the curvature of the risk from derailing convergence.

Recommendations

• Nonparametric hazard estimation with time-varying discrete covariates
• Nonparametric survival analysis with time-dependent covariate effects: A penalized partial likelihood approach
• Estimating multiplicative and additive hazard functions by kernel methods
• A gradient boosting algorithm for survival analysis via direct optimization of concordance index
• Boosting method for nonlinear transformation models with censored survival data

Cites work

• 10.1162/1532443041424319
• Additive modelling of functional gradients
• Boosting With theL₂Loss
• Boosting additive models using component-wise P-splines
• Boosting algorithms: regularization, prediction and model fitting
• Boosting with early stopping: convergence and consistency
• Characterizing \(L_{2}\)Boosting
• First-order complete and computationally complete query languages for spatio-temporal databases
• Greedy function approximation: A gradient boosting machine.
• Introductory lectures on convex optimization. A basic course.
• Likelihood Ratio Tests for Model Selection and Non-Nested Hypotheses
• Nonparametric inference for a family of counting processes
• On the ``degrees of freedom of the lasso
• Population theory for boosting ensembles.
• Prediction Games and Arcing Algorithms
• Regularized Estimation in the Accelerated Failure Time Model with High-Dimensional Covariates
• Survival ensembles
• The L₂ Rate of Convergence for Event History Regression with Time‐dependent Covariates
• Transformation boosting machines
• Weak convergence and empirical processes. With applications to statistics

Cited in

(4)

• Unbiased Boosting Estimation for Censored Survival Data
• Regularisation in discrete survival models: a comparison of Lasso and gradient boosting
• Accelerated failure time models with error-prone response and nonlinear covariates
• Boosted coefficient models

This page was built for publication: Boosted nonparametric hazards with time-dependent covariates

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2054480)