Tree-structured prognostic classification for censored survival data: Validation of computationally inexpensive model selection criteria

Abdissa Negassa, Antonio Ciampi, Michal Abrahamowicz, Stanley Shapiro, Jean François Boivin

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

The performance of computationally inexpensive model selection criteria in the context of tree structured prediction is discussed. It is shown through a simulation study that no one model selection criterion exhibits a uniformly superior performance over a wide range of scenarios. Therefore, a two-stage approach for model selection is suggested and shown to perform satisfactorily. A computationally efficient method of tree-growing within the RECursive Partition and AMalgamation (RECPAM) framework is suggested. The computationally efficient algorithm gives identical results as the original RECPAM tree-growing algorithm. An example of medical data analysis for developing prognostic classification is presented.

Original languageEnglish (US)
Pages (from-to)289-317
Number of pages29
JournalJournal of Statistical Computation and Simulation
Volume67
Issue number4
StatePublished - 2000
Externally publishedYes

Fingerprint

Censored Survival Data
Model Selection Criteria
Amalgamation
Partition
Trees (mathematics)
Model Selection
Data analysis
Efficient Algorithms
Simulation Study
Scenarios
Prediction
Range of data
Model selection criteria

Keywords

  • Censored survival data
  • Prognostic classification
  • RECPAM
  • Regression trees
  • Two-stage approach

ASJC Scopus subject areas

  • Applied Mathematics
  • Modeling and Simulation
  • Statistics and Probability

Cite this

Tree-structured prognostic classification for censored survival data : Validation of computationally inexpensive model selection criteria. / Negassa, Abdissa; Ciampi, Antonio; Abrahamowicz, Michal; Shapiro, Stanley; Boivin, Jean François.

In: Journal of Statistical Computation and Simulation, Vol. 67, No. 4, 2000, p. 289-317.

Research output: Contribution to journalArticle

@article{d6b8e09827c7447ea01b253395c16d60,
title = "Tree-structured prognostic classification for censored survival data: Validation of computationally inexpensive model selection criteria",
abstract = "The performance of computationally inexpensive model selection criteria in the context of tree structured prediction is discussed. It is shown through a simulation study that no one model selection criterion exhibits a uniformly superior performance over a wide range of scenarios. Therefore, a two-stage approach for model selection is suggested and shown to perform satisfactorily. A computationally efficient method of tree-growing within the RECursive Partition and AMalgamation (RECPAM) framework is suggested. The computationally efficient algorithm gives identical results as the original RECPAM tree-growing algorithm. An example of medical data analysis for developing prognostic classification is presented.",
keywords = "Censored survival data, Prognostic classification, RECPAM, Regression trees, Two-stage approach",
author = "Abdissa Negassa and Antonio Ciampi and Michal Abrahamowicz and Stanley Shapiro and Boivin, {Jean Fran{\cc}ois}",
year = "2000",
language = "English (US)",
volume = "67",
pages = "289--317",
journal = "Journal of Statistical Computation and Simulation",
issn = "0094-9655",
publisher = "Taylor and Francis Ltd.",
number = "4",

}

TY - JOUR

T1 - Tree-structured prognostic classification for censored survival data

T2 - Validation of computationally inexpensive model selection criteria

AU - Negassa, Abdissa

AU - Ciampi, Antonio

AU - Abrahamowicz, Michal

AU - Shapiro, Stanley

AU - Boivin, Jean François

PY - 2000

Y1 - 2000

N2 - The performance of computationally inexpensive model selection criteria in the context of tree structured prediction is discussed. It is shown through a simulation study that no one model selection criterion exhibits a uniformly superior performance over a wide range of scenarios. Therefore, a two-stage approach for model selection is suggested and shown to perform satisfactorily. A computationally efficient method of tree-growing within the RECursive Partition and AMalgamation (RECPAM) framework is suggested. The computationally efficient algorithm gives identical results as the original RECPAM tree-growing algorithm. An example of medical data analysis for developing prognostic classification is presented.

AB - The performance of computationally inexpensive model selection criteria in the context of tree structured prediction is discussed. It is shown through a simulation study that no one model selection criterion exhibits a uniformly superior performance over a wide range of scenarios. Therefore, a two-stage approach for model selection is suggested and shown to perform satisfactorily. A computationally efficient method of tree-growing within the RECursive Partition and AMalgamation (RECPAM) framework is suggested. The computationally efficient algorithm gives identical results as the original RECPAM tree-growing algorithm. An example of medical data analysis for developing prognostic classification is presented.

KW - Censored survival data

KW - Prognostic classification

KW - RECPAM

KW - Regression trees

KW - Two-stage approach

UR - http://www.scopus.com/inward/record.url?scp=0034349502&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0034349502&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0034349502

VL - 67

SP - 289

EP - 317

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

SN - 0094-9655

IS - 4

ER -