Nonsmooth DC programming approach to clusterwise linear regression: optimality conditions and algorithms. (2nd January 2018)

• Record Type: Journal Article
• Authors: Bagirov, A.M.; Ugon, J.
• Is Part Of: Optimization methods and software. Volume 33:Number 1(2018)
• Page Start: 194
• Page End: 219
• Publication Date: 2018-01-02
• Subjects: nonsmooth optimization -- DC programming -- regression analysis -- cluster analysis; 90C26 -- 90C56; Mathematical optimization -- Periodicals; Algorithms -- Periodicals; 519.7
• DOI: 10.1080/10556788.2017.1371717
• Languages: English
• ISSNs: 1055-6788

Abstract

Abstract : 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.