Minimization of Akaike's information criterion in linear regression analysis via mixed integer nonlinear program

DOI10.1080/10556788.2017.1333611zbMATH Open1398.90111arXiv1606.05030OpenAlexW3098332525MaRDI QIDQ4637833FDOQ4637833

Authors: Keiji Kimura, Hayato Waki 

Publication date: 3 May 2018

Published in: Optimization Methods \& Software (Search for Journal in Brave)

Abstract: Akaike's information criterion (AIC) is a measure of the quality of a statistical model for a given set of data. We can determine the best statistical model for a particular data set by the minimization of the AIC. Since we need to evaluate exponentially many candidates of the model by the minimization of the AIC, the minimization is unreasonable. Instead, stepwise methods, which are local search algorithms, are commonly used to find a better statistical model though it may not be the best. We propose a branch and bound search algorithm for a mixed integer nonlinear programming formulation of the AIC minimization by Miyashiro and Takano (2015). More concretely, we propose methods to find lower and upper bounds, and branching rules for this minimization. We then combine them with SCIP, which is a mathematical optimization software and a branch-and-bound framework. We show that the proposed method can provide the best statistical model based on AIC for small-sized or medium-sized benchmark data sets in UCI Machine Learning Repository. Furthermore, we show that this method finds good quality solutions for large-sized benchmark data sets.

Full work available at URL: https://arxiv.org/abs/1606.05030

Recommendations

• Mixed integer nonlinear program for minimization of Akaike's information criterion
• Application of a mixed integer nonlinear programming approach to variable selection in logistic regression
• Extending AIC to best subset regression
• Information criteria and statistical modeling.
• An application of multiple comparison techniques to model selection

zbMATH Keywords

branch-and-boundmixed integer nonlinear programSCIP and Akaike's information criterion

Mathematics Subject Classification ID

Linear regression; mixed models (62J05) Polyhedral combinatorics, branch-and-bound, branch-and-cut (90C57) Mixed integer programming (90C11)

Cites Work

• LAPACK Users' Guide
• SCIP: solving constraint integer programs
• Title not available (Why is that?)
• 10.1162/153244303322753616
• A new look at the statistical model identification
• Best subset selection via a modern optimization lens
• Algorithm for cardinality-constrained quadratic optimization
• Mixed integer second-order cone programming formulations for variable selection in linear regression
• SCIP: global optimization of mixed-integer nonlinear programs in a branch-and-cut framework
• Constraint Pairing In Integer Programming
• Mixed integer nonlinear program for minimization of Akaike's information criterion

Cited In (7)

• A Mixed-Integer Fractional Optimization Approach to Best Subset Selection
• Mixed integer nonlinear program for minimization of Akaike's information criterion
• Title not available (Why is that?)
• Application of a mixed integer nonlinear programming approach to variable selection in logistic regression
• An efficient optimization approach for best subset selection in linear regression, with application to model selection and fitting in autoregressive time-series
• Mixed integer quadratic optimization formulations for eliminating multicollinearity based on variance inflation factor
• Best subset selection via cross-validation criterion

Uses Software

• LAPACK
• R
• SCIP
• UG
• UCI-ml
• CPLEX
• ParaSCIP
• MAIC

This page was built for publication: Minimization of Akaike's information criterion in linear regression analysis via mixed integer nonlinear program

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4637833)