View : 724 Download: 0
SIMPLIFIED COPOSITIVE AND LAGRANGIAN RELAXATIONS FOR LINEARLY CONSTRAINED QUADRATIC OPTIMIZATION PROBLEMS IN CONTINUOUS AND BINARY VARIABLES
- Title
- SIMPLIFIED COPOSITIVE AND LAGRANGIAN RELAXATIONS FOR LINEARLY CONSTRAINED QUADRATIC OPTIMIZATION PROBLEMS IN CONTINUOUS AND BINARY VARIABLES
- Authors
- Arima, Naohiko; Kim, Sunyoung; Kojima, Masakazu
- Ewha Authors
- 김선영
- SCOPUS Author ID
- 김선영
- Issue Date
- 2014
- Journal Title
- PACIFIC JOURNAL OF OPTIMIZATION
- ISSN
- 1348-9151
- Citation
- PACIFIC JOURNAL OF OPTIMIZATION vol. 10, no. 3, pp. 437 - 451
- Keywords
- nonconvex quadratic optimization; 0-1 mixed integer program; Lagrangian relaxation; copositve programming relaxation; completely positive programming relaxation
- Publisher
- YOKOHAMA PUBL
- Indexed
- SCIE
- Document Type
- Article
- Abstract
- For a quadratic optimization problem (QOP) with linear equality constraints in continuous non-negative variables and binary variables, vie propose three relaxations in simplified forms with a parameter lambda: Lagrangian, completely positive, and copositive relaxations. These relaxations are obtained by reducing the QOP to an equivalent QOP with a single quadratic equality constraint in nonnegative variables, and applying the Lagrangian relaxation to the resulting QOP. As a result, an unconstrained QOP with a Lagrangian multiplier lambda in nonnegative variables is obtained. The other two relaxations are a primal-dual pair of a completely positive programming (CPP) relaxation in a variable matrix with the upper-left element set to 1 and a copositive programming (CP) relaxation in a single variable. The CPP relaxation is derived from the unconstrained QOP with the parameter lambda, based on the recent result by Arima, Kim and Kojima. The three relaxations with a same parameter value lambda > 0 work as relaxations of the original QOP. The optimal values zeta(lambda) of the three relaxations coincide, and monotonically converge to the optimal value of the original QOP as lambda tends to infinity under a moderate assumption. The parameter lambda serves as a penalty parameter when it is chosen to be positive. Thus, the standard theory on the penalty function method can be applied to establish the convergence.
- Appears in Collections:
- 자연과학대학 > 수학전공 > Journal papers
- Files in This Item:
There are no files associated with this item.
- Export
- RIS (EndNote)
- XLS (Excel)
- XML