Deakin University
Browse

Nonsmooth DC programming approach to clusterwise linear regression: optimality conditions and algorithms

Version 2 2024-06-04, 13:49
Version 1 2018-04-16, 16:03
journal contribution
posted on 2018-01-01, 00:00 authored by A M Bagirov, Julien UgonJulien Ugon
© 2017 Informa UK Limited, trading as Taylor & Francis Group. The clusterwise linear regression problem is formulated as a nonsmooth nonconvex optimization problem using the squared regression error function. The objective function in this problem is represented as a difference of convex functions. Optimality conditions are derived, and an algorithm is designed based on such a representation. An incremental approach is proposed to generate starting solutions. The algorithm is tested on small to large data sets.

History

Journal

Optimization methods and software

Volume

33

Issue

1

Pagination

194 - 219

Publisher

Taylor & Francis

Location

Abingdon, Eng.

ISSN

1055-6788

eISSN

1029-4937

Language

eng

Publication classification

C Journal article; C1.1 Refereed article in a scholarly journal

Copyright notice

2017, Informa UK Limited