Bagging and deep learning in optimal individualized treatment rules

Xinlei Mi, Fei Zou, Ruoqing Zhu

Research output: Contribution to journalArticle

Abstract

An ENsemble Deep Learning Optimal Treatment (EndLot) approach is proposed for personalized medicine problems. The statistical framework of the proposed method is based on the outcome weighted learning (OWL) framework which transforms the optimal decision rule problem into a weighted classification problem. We further employ an ensemble of deep neural networks (DNNs) to learn the optimal decision rule. Utilizing the flexibility of DNNs and the stability of bootstrap aggregation, the proposed method achieves a considerable improvement over existing methods. An R package “ITRlearn” is developed to implement the proposed method. Numerical performance is demonstrated via simulation studies and a real data analysis of the Cancer Cell Line Encyclopedia data.

Original languageEnglish (US)
JournalBiometrics
DOIs
StatePublished - Jan 1 2019

Fingerprint

Bagging
learning
Learning
Decision Rules
neural networks
Medicine
Agglomeration
Cells
Neural Networks
Encyclopedias
Ensemble Learning
Precision Medicine
Therapeutics
methodology
Classification Problems
Bootstrap
Aggregation
data analysis
Data analysis
Cancer

Keywords

  • Bootstrap aggregating
  • deep neural network
  • high-dimensional data
  • outcome weighted learning
  • personalized medicine

ASJC Scopus subject areas

  • Statistics and Probability
  • Biochemistry, Genetics and Molecular Biology(all)
  • Immunology and Microbiology(all)
  • Agricultural and Biological Sciences(all)
  • Applied Mathematics

Cite this

Bagging and deep learning in optimal individualized treatment rules. / Mi, Xinlei; Zou, Fei; Zhu, Ruoqing.

In: Biometrics, 01.01.2019.

Research output: Contribution to journalArticle

@article{a3c1908577ef47d2b4e1806c734d8082,
title = "Bagging and deep learning in optimal individualized treatment rules",
abstract = "An ENsemble Deep Learning Optimal Treatment (EndLot) approach is proposed for personalized medicine problems. The statistical framework of the proposed method is based on the outcome weighted learning (OWL) framework which transforms the optimal decision rule problem into a weighted classification problem. We further employ an ensemble of deep neural networks (DNNs) to learn the optimal decision rule. Utilizing the flexibility of DNNs and the stability of bootstrap aggregation, the proposed method achieves a considerable improvement over existing methods. An R package “ITRlearn” is developed to implement the proposed method. Numerical performance is demonstrated via simulation studies and a real data analysis of the Cancer Cell Line Encyclopedia data.",
keywords = "Bootstrap aggregating, deep neural network, high-dimensional data, outcome weighted learning, personalized medicine",
author = "Xinlei Mi and Fei Zou and Ruoqing Zhu",
year = "2019",
month = "1",
day = "1",
doi = "10.1111/biom.12990",
language = "English (US)",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",

}

TY - JOUR

T1 - Bagging and deep learning in optimal individualized treatment rules

AU - Mi, Xinlei

AU - Zou, Fei

AU - Zhu, Ruoqing

PY - 2019/1/1

Y1 - 2019/1/1

N2 - An ENsemble Deep Learning Optimal Treatment (EndLot) approach is proposed for personalized medicine problems. The statistical framework of the proposed method is based on the outcome weighted learning (OWL) framework which transforms the optimal decision rule problem into a weighted classification problem. We further employ an ensemble of deep neural networks (DNNs) to learn the optimal decision rule. Utilizing the flexibility of DNNs and the stability of bootstrap aggregation, the proposed method achieves a considerable improvement over existing methods. An R package “ITRlearn” is developed to implement the proposed method. Numerical performance is demonstrated via simulation studies and a real data analysis of the Cancer Cell Line Encyclopedia data.

AB - An ENsemble Deep Learning Optimal Treatment (EndLot) approach is proposed for personalized medicine problems. The statistical framework of the proposed method is based on the outcome weighted learning (OWL) framework which transforms the optimal decision rule problem into a weighted classification problem. We further employ an ensemble of deep neural networks (DNNs) to learn the optimal decision rule. Utilizing the flexibility of DNNs and the stability of bootstrap aggregation, the proposed method achieves a considerable improvement over existing methods. An R package “ITRlearn” is developed to implement the proposed method. Numerical performance is demonstrated via simulation studies and a real data analysis of the Cancer Cell Line Encyclopedia data.

KW - Bootstrap aggregating

KW - deep neural network

KW - high-dimensional data

KW - outcome weighted learning

KW - personalized medicine

UR - http://www.scopus.com/inward/record.url?scp=85063611450&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85063611450&partnerID=8YFLogxK

U2 - 10.1111/biom.12990

DO - 10.1111/biom.12990

M3 - Article

JO - Biometrics

JF - Biometrics

SN - 0006-341X

ER -