Stabilizing high-dimensional prediction models using feature graphs

Shivapratap Gopakumar, Truyen Tran, Tu Dinh Nguyen, Dinh Phung, Svetha Venkatesh

Research output: Contribution to journalArticleResearchpeer-review

6 Citations (Scopus)

Abstract

We investigate feature stability in the context of clinical prognosis derived from high-dimensional electronic medical records. To reduce variance in the selected features that are predictive, we introduce Laplacian-based regularization into a regression model. The Laplacian is derived on a feature graph that captures both the temporal and hierarchic relations between hospital events, diseases, and interventions. Using a cohort of patients with heart failure, we demonstrate better feature stability and goodness-of-fit through feature graph stabilization.

Original languageEnglish
Article number6887285
Pages (from-to)1044-1052
Number of pages9
JournalIEEE Journal of Biomedical and Health Informatics
Volume19
Issue number3
DOIs
Publication statusPublished - May 2015
Externally publishedYes

Cite this

Gopakumar, Shivapratap ; Tran, Truyen ; Nguyen, Tu Dinh ; Phung, Dinh ; Venkatesh, Svetha. / Stabilizing high-dimensional prediction models using feature graphs. In: IEEE Journal of Biomedical and Health Informatics. 2015 ; Vol. 19, No. 3. pp. 1044-1052.
@article{9131fb4dff9d43e4b6cf076142e15a4d,
title = "Stabilizing high-dimensional prediction models using feature graphs",
abstract = "We investigate feature stability in the context of clinical prognosis derived from high-dimensional electronic medical records. To reduce variance in the selected features that are predictive, we introduce Laplacian-based regularization into a regression model. The Laplacian is derived on a feature graph that captures both the temporal and hierarchic relations between hospital events, diseases, and interventions. Using a cohort of patients with heart failure, we demonstrate better feature stability and goodness-of-fit through feature graph stabilization.",
author = "Shivapratap Gopakumar and Truyen Tran and Nguyen, {Tu Dinh} and Dinh Phung and Svetha Venkatesh",
year = "2015",
month = "5",
doi = "10.1109/JBHI.2014.2353031",
language = "English",
volume = "19",
pages = "1044--1052",
journal = "IEEE Journal of Biomedical and Health Informatics",
issn = "2168-2194",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
number = "3",

}

Stabilizing high-dimensional prediction models using feature graphs. / Gopakumar, Shivapratap; Tran, Truyen; Nguyen, Tu Dinh; Phung, Dinh; Venkatesh, Svetha.

In: IEEE Journal of Biomedical and Health Informatics, Vol. 19, No. 3, 6887285, 05.2015, p. 1044-1052.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Stabilizing high-dimensional prediction models using feature graphs

AU - Gopakumar, Shivapratap

AU - Tran, Truyen

AU - Nguyen, Tu Dinh

AU - Phung, Dinh

AU - Venkatesh, Svetha

PY - 2015/5

Y1 - 2015/5

N2 - We investigate feature stability in the context of clinical prognosis derived from high-dimensional electronic medical records. To reduce variance in the selected features that are predictive, we introduce Laplacian-based regularization into a regression model. The Laplacian is derived on a feature graph that captures both the temporal and hierarchic relations between hospital events, diseases, and interventions. Using a cohort of patients with heart failure, we demonstrate better feature stability and goodness-of-fit through feature graph stabilization.

AB - We investigate feature stability in the context of clinical prognosis derived from high-dimensional electronic medical records. To reduce variance in the selected features that are predictive, we introduce Laplacian-based regularization into a regression model. The Laplacian is derived on a feature graph that captures both the temporal and hierarchic relations between hospital events, diseases, and interventions. Using a cohort of patients with heart failure, we demonstrate better feature stability and goodness-of-fit through feature graph stabilization.

UR - http://www.scopus.com/inward/record.url?scp=84929376633&partnerID=8YFLogxK

U2 - 10.1109/JBHI.2014.2353031

DO - 10.1109/JBHI.2014.2353031

M3 - Article

VL - 19

SP - 1044

EP - 1052

JO - IEEE Journal of Biomedical and Health Informatics

JF - IEEE Journal of Biomedical and Health Informatics

SN - 2168-2194

IS - 3

M1 - 6887285

ER -