Hierarchical subspace models for contingency tables

Hisayuki Hara, Tomonari Sei, A. Takemura Akimichi

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

For the statistical analysis of multiway contingency tables, we propose modeling interaction terms in each maximal compact component of a hierarchical model. By this approach we can search for parsimonious models with smaller degrees of freedom than the usual hierarchical model, while preserving the localization property of the inference in the hierarchical model. This approach also enables us to evaluate the localization property of a given log-affine model. We discuss estimation and exact tests of the proposed model and illustrate the advantage of the proposed modeling with some data sets.

Original languageEnglish
Pages (from-to)19-34
Number of pages16
JournalJournal of Multivariate Analysis
Volume103
Issue number1
DOIs
Publication statusPublished - 2012 Jan

Fingerprint

Contingency Table
Hierarchical Model
Subspace
Exact Test
Modeling
Statistical Analysis
Degree of freedom
Model
Evaluate
Term
Interaction
Hierarchical model
Contingency table
Statistical methods
Localization

Keywords

  • 62H05
  • 62H17
  • Context specific interaction model
  • Divider
  • Markov bases
  • Split model
  • Uniform association model

ASJC Scopus subject areas

  • Statistics, Probability and Uncertainty
  • Numerical Analysis
  • Statistics and Probability

Cite this

Hierarchical subspace models for contingency tables. / Hara, Hisayuki; Sei, Tomonari; Takemura Akimichi, A.

In: Journal of Multivariate Analysis, Vol. 103, No. 1, 01.2012, p. 19-34.

Research output: Contribution to journalArticle

Hara, Hisayuki ; Sei, Tomonari ; Takemura Akimichi, A. / Hierarchical subspace models for contingency tables. In: Journal of Multivariate Analysis. 2012 ; Vol. 103, No. 1. pp. 19-34.
@article{83c3e791ca664ba19c9e91f5d83d57a4,
title = "Hierarchical subspace models for contingency tables",
abstract = "For the statistical analysis of multiway contingency tables, we propose modeling interaction terms in each maximal compact component of a hierarchical model. By this approach we can search for parsimonious models with smaller degrees of freedom than the usual hierarchical model, while preserving the localization property of the inference in the hierarchical model. This approach also enables us to evaluate the localization property of a given log-affine model. We discuss estimation and exact tests of the proposed model and illustrate the advantage of the proposed modeling with some data sets.",
keywords = "62H05, 62H17, Context specific interaction model, Divider, Markov bases, Split model, Uniform association model",
author = "Hisayuki Hara and Tomonari Sei and {Takemura Akimichi}, A.",
year = "2012",
month = "1",
doi = "10.1016/j.jmva.2011.06.003",
language = "English",
volume = "103",
pages = "19--34",
journal = "Journal of Multivariate Analysis",
issn = "0047-259X",
publisher = "Academic Press Inc.",
number = "1",

}

TY - JOUR

T1 - Hierarchical subspace models for contingency tables

AU - Hara, Hisayuki

AU - Sei, Tomonari

AU - Takemura Akimichi, A.

PY - 2012/1

Y1 - 2012/1

N2 - For the statistical analysis of multiway contingency tables, we propose modeling interaction terms in each maximal compact component of a hierarchical model. By this approach we can search for parsimonious models with smaller degrees of freedom than the usual hierarchical model, while preserving the localization property of the inference in the hierarchical model. This approach also enables us to evaluate the localization property of a given log-affine model. We discuss estimation and exact tests of the proposed model and illustrate the advantage of the proposed modeling with some data sets.

AB - For the statistical analysis of multiway contingency tables, we propose modeling interaction terms in each maximal compact component of a hierarchical model. By this approach we can search for parsimonious models with smaller degrees of freedom than the usual hierarchical model, while preserving the localization property of the inference in the hierarchical model. This approach also enables us to evaluate the localization property of a given log-affine model. We discuss estimation and exact tests of the proposed model and illustrate the advantage of the proposed modeling with some data sets.

KW - 62H05

KW - 62H17

KW - Context specific interaction model

KW - Divider

KW - Markov bases

KW - Split model

KW - Uniform association model

UR - http://www.scopus.com/inward/record.url?scp=79960889496&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79960889496&partnerID=8YFLogxK

U2 - 10.1016/j.jmva.2011.06.003

DO - 10.1016/j.jmva.2011.06.003

M3 - Article

AN - SCOPUS:79960889496

VL - 103

SP - 19

EP - 34

JO - Journal of Multivariate Analysis

JF - Journal of Multivariate Analysis

SN - 0047-259X

IS - 1

ER -