Tree based weighted learning for estimating individualized treatment rules with censored data

DOI10.1214/17-EJS1305MaRDI QIDQ2412261zbMATH OpenOpenAlexWikidataFDO

Authors Yifan Cui, Ruoqing Zhu, Michael R. Kosorok

Publication date 23 October 2017

Published in Electronic Journal of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1707.09632, https://projecteuclid.org/euclid.ejs/1508292531

zbMATH Keywords

consistencyindividualized treatment rulenonparametric estimationoutcome weighted learningpersonalized medicinerecursively imputed survival treesright censored data

Mathematics Subject Classification ID

Nonparametric estimation (62G05) Asymptotic properties of nonparametric inference (62G20) Censored data models (62N01) Applications of statistics to biology and medical sciences; meta analysis (62P10) Learning and adaptive systems in artificial intelligence (68T05)

Abstract: Estimating individualized treatment rules is a central task for personalized medicine. [zhao2012estimating] and [zhang2012robust] proposed outcome weighted learning to estimate individualized treatment rules directly through maximizing the expected outcome without modeling the response directly. In this paper, we extend the outcome weighted learning to right censored survival data without requiring either an inverse probability of censoring weighting or a semiparametric modeling of the censoring and failure times as done in [zhao2015doubly]. To accomplish this, we take advantage of the tree based approach proposed in [zhu2012recursively] to nonparametrically impute the survival time in two different ways. The first approach replaces the reward of each individual by the expected survival time, while in the second approach only the censored observations are imputed by their conditional expected failure times. We establish consistency and convergence rates for both estimators. In simulation studies, our estimators demonstrate improved performance compared to existing methods. We also illustrate the proposed method on a phase III clinical trial of non-small cell lung cancer.

Review text

In this paper, the authors investigate a major problem in the field of personalized medicine. Personalized medicine is an emerging research area and consists, broadly speaking, in the estimation of the correct treatment and/or the correct dose as a function of the patient's covariates. In particular, this paper is focused on the estimation of the survival function in the presence of right-censored data. In such a setting, the response of interest is only partially observed. To overcome this problem, a nonparametric tree-based approach is introduced. The problem is that the survival time is not available for all the subjects and must be estimated. Two different estimation procedures are defined: a first one considers the estimation of the survival time for all the subjects, while the second one uses the estimated values only in the case of censored times. Under mild conditions, the authors prove the consistency of the method, and a result about the rate of convergence is also provided. Finally, the authors show that the method proposed here outperforms the classical competitors. This is done through a simulation study with 4 different scenarios.

Recommendations

• Doubly robust learning for estimating individualized treatment with censored data
• Greedy outcome weighted tree learning of optimal personalized treatment rules
• Tree-based multivariate regression and density estimation with right-censored data
• Estimating individualized treatment rules using outcome weighted learning
• Tree-based methods for individualized treatment regimes

Cited in

(28)

• D-learning to estimate optimal individual treatment rules
• A modified classification tree method for personalized medicine decisions
• Ensemble survival trees for identifying subpopulations in personalized medicine
• Data-guided Treatment Recommendation with Feature Scores
• Consistency of survival tree and forest models: splitting bias and correction
• Multi-Armed Angle-Based Direct Learning for Estimating Optimal Individualized Treatment Rules With Various Outcomes
• scientific article; zbMATH DE number 7306879 (Why is no real title available?)
• Randomization as regularization: a degrees of freedom explanation for random forest success
• Doubly robust learning for estimating individualized treatment with censored data
• Greedy outcome weighted tree learning of optimal personalized treatment rules
• A semiparametric instrumental variable approach to optimal treatment regimes under endogeneity
• Multicategory Angle-Based Learning for Estimating Optimal Dynamic Treatment Regimes With Censored Data
• The Impact of Churn on Client Value in Health Insurance, Evaluation Using a Random Forest Under Various Censoring Mechanisms
• Resampling‐based confidence intervals for model‐free robust inference on optimal treatment regimes
• On Robustness of Individualized Decision Rules
• Estimating Tree-Based Dynamic Treatment Regimes Using Observational Data with Restricted Treatment Sequences
• Transformation-Invariant Learning of Optimal Individualized Decision Rules with Time-to-Event Outcomes
• Doubly robust estimation of optimal treatment regimes for survival data using an instrumental variable
• Selection of the optimal personalized treatment from multiple treatments with right-censored multivariate outcome measures
• Stabilized direct learning for efficient estimation of individualized treatment rules
• Survival analysis using deep learning with medical imaging
• Estimation in optimal treatment regimes based on mean residual lifetimes with right-censored data
• Estimating individual treatment effects by gradient boosting trees
• A nonparametric method for value function guided subgroup identification via gradient tree boosting for censored survival data
• M-learning for individual treatment rule with survival outcomes
• Efficient and robust transfer learning of optimal individualized treatment regimes with right-censored survival data
• Smoothed estimation on optimal treatment regime under semisupervised setting in randomized trials
• A rank-based approach to estimating monotone individualized two treatment regimes

This page was built for publication: Tree based weighted learning for estimating individualized treatment rules with censored data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2412261)