A new approach to select the best subset of predictors in linear regression modelling: bi-objective mixed integer linear programming

Title: A new approach to select the best subset of predictors in linear regression modelling: bi-objective mixed integer linear programming
Creator: Charkhgard, Hadi; Eshragh, Ali
Relation: ANZIAM Journal Vol. 61, Issue 1, p. 64-75
Publisher Link: http://dx.doi.org/10.1017/S1446181118000275
Publisher: Cambridge University Press
Resource Type: journal article
Date: 2019
Description: We study the problem of choosing the best subset of features in linear regression, given observations. This problem naturally contains two objective functions including minimizing the amount of bias and minimizing the number of predictors. The existing approaches transform the problem into a single-objective optimization problem. We explain the main weaknesses of existing approaches and, to overcome their drawbacks, we propose a bi-objective mixed integer linear programming approach. A computational study shows the efficacy of the proposed approach.
Subject: linear regression; best subset selection; bi-objective mixed integer linear programming
Identifier: http://hdl.handle.net/1959.13/1400977
Identifier: uon:34844
Identifier: ISSN:1446-1811
Language: eng
Reviewed

		Thumbnail	File	Description	Size	Format