An efficient branch-and-bound strategy for subset vector autoregressive model selection

Cristian Gatu (Corresponding Author), Erricos J. Kontoghiorghes, Manfred Gilli, Peter Winker

Research output: Contribution to journalArticleScientificpeer-review

10 Citations (Scopus)

Abstract

A computationally efficient branch-and-bound strategy for finding the subsets of the most statistically significant variables of a vector autoregressive (VAR) model from a given search subspace is proposed. Specifically, the candidate submodels are obtained by deleting columns from the coefficient matrices of the full-specified VAR process. The strategy is based on a regression tree and derives the best-subset VAR models without computing the whole tree. The branch-and-bound cutting test is based on monotone statistical selection criteria which are functions of the determinant of the estimated residual covariance matrix. Experimental results confirm the computational efficiency of the proposed algorithm.
Original languageEnglish
Pages (from-to)1949-1963
JournalJournal of Economic Dynamics and Control
Volume32
Issue number6
DOIs
Publication statusPublished - 2008
MoE publication typeA1 Journal article-refereed

Fingerprint

Vector Autoregressive Model
Branch-and-bound
Model Selection
Vector Autoregressive Process
Regression Tree
Subset
Computational Efficiency
Covariance matrix
Monotone
Determinant
Subspace
Computational efficiency
Computing
Experimental Results
Coefficient
Strategy
Vector autoregressive model
Model selection
Vector autoregressive process
Regression tree

Keywords

  • vector autoregressive model
  • model selection
  • branch-and-bound algorithms

Cite this

Gatu, Cristian ; Kontoghiorghes, Erricos J. ; Gilli, Manfred ; Winker, Peter. / An efficient branch-and-bound strategy for subset vector autoregressive model selection. In: Journal of Economic Dynamics and Control. 2008 ; Vol. 32, No. 6. pp. 1949-1963.
@article{b47ad9479655406494f76a01ecf275c9,
title = "An efficient branch-and-bound strategy for subset vector autoregressive model selection",
abstract = "A computationally efficient branch-and-bound strategy for finding the subsets of the most statistically significant variables of a vector autoregressive (VAR) model from a given search subspace is proposed. Specifically, the candidate submodels are obtained by deleting columns from the coefficient matrices of the full-specified VAR process. The strategy is based on a regression tree and derives the best-subset VAR models without computing the whole tree. The branch-and-bound cutting test is based on monotone statistical selection criteria which are functions of the determinant of the estimated residual covariance matrix. Experimental results confirm the computational efficiency of the proposed algorithm.",
keywords = "vector autoregressive model, model selection, branch-and-bound algorithms",
author = "Cristian Gatu and Kontoghiorghes, {Erricos J.} and Manfred Gilli and Peter Winker",
year = "2008",
doi = "10.1016/j.jedc.2007.08.001",
language = "English",
volume = "32",
pages = "1949--1963",
journal = "Journal of Economic Dynamics and Control",
issn = "0165-1889",
publisher = "Elsevier",
number = "6",

}

An efficient branch-and-bound strategy for subset vector autoregressive model selection. / Gatu, Cristian (Corresponding Author); Kontoghiorghes, Erricos J.; Gilli, Manfred; Winker, Peter.

In: Journal of Economic Dynamics and Control, Vol. 32, No. 6, 2008, p. 1949-1963.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - An efficient branch-and-bound strategy for subset vector autoregressive model selection

AU - Gatu, Cristian

AU - Kontoghiorghes, Erricos J.

AU - Gilli, Manfred

AU - Winker, Peter

PY - 2008

Y1 - 2008

N2 - A computationally efficient branch-and-bound strategy for finding the subsets of the most statistically significant variables of a vector autoregressive (VAR) model from a given search subspace is proposed. Specifically, the candidate submodels are obtained by deleting columns from the coefficient matrices of the full-specified VAR process. The strategy is based on a regression tree and derives the best-subset VAR models without computing the whole tree. The branch-and-bound cutting test is based on monotone statistical selection criteria which are functions of the determinant of the estimated residual covariance matrix. Experimental results confirm the computational efficiency of the proposed algorithm.

AB - A computationally efficient branch-and-bound strategy for finding the subsets of the most statistically significant variables of a vector autoregressive (VAR) model from a given search subspace is proposed. Specifically, the candidate submodels are obtained by deleting columns from the coefficient matrices of the full-specified VAR process. The strategy is based on a regression tree and derives the best-subset VAR models without computing the whole tree. The branch-and-bound cutting test is based on monotone statistical selection criteria which are functions of the determinant of the estimated residual covariance matrix. Experimental results confirm the computational efficiency of the proposed algorithm.

KW - vector autoregressive model

KW - model selection

KW - branch-and-bound algorithms

U2 - 10.1016/j.jedc.2007.08.001

DO - 10.1016/j.jedc.2007.08.001

M3 - Article

VL - 32

SP - 1949

EP - 1963

JO - Journal of Economic Dynamics and Control

JF - Journal of Economic Dynamics and Control

SN - 0165-1889

IS - 6

ER -