- Research Article
- Open Access
-Duality Theorems for Convex Semidefinite Optimization Problems with Conic Constraints
Journal of Inequalities and Applications volume 2010, Article number: 363012 (2010)
A convex semidefinite optimization problem with a conic constraint is considered. We formulate a Wolfe-type dual problem for the problem for its -approximate solutions, and then we prove -weak duality theorem and -strong duality theorem which hold between the problem and its Wolfe type dual problem. Moreover, we give an example illustrating the duality theorems.
Convex semidefinite optimization problem is to optimize an objective convex function over a linear matrix inequality. When the objective function is linear and the corresponding matrices are diagonal, this problem becomes a linear optimization problem.
For convex semidefinite optimization problem, Lagrangean duality without constraint qualification [1, 2], complete dual characterization conditions of solutions [1, 3, 4], saddle point theorems , and characterizations of optimal solution sets [6, 7] have been investigated.
Recently, Jeyakumar and Glover  gave -optimality conditions for convex optimization problems, which hold without any constraint qualification. Yokoyama and Shiraishi  gave a special case of convex optimization problem which satisfies -optimality conditions. Kim and Lee  proved sequential -saddle point theorems and -duality theorems for convex semidefinite optimization problems which have not conic constraints.
The purpose of this paper is to extend the -duality theorems by Kim and Lee  to convex semidefinite optimization problems with conic constraints. We formulate a Wolfe type dual problem for the problem for its -approximate solutions, and then prove -weak duality theorem and -strong duality theorem for the problem and its Wolfe type dual problem, which hold under a weakened constraint qualification. Moreover, we give an example illustrating the duality theorems.
Consider the following convex semidefinite optimization problem:
where is a convex function, is a closed convex cone of , and for , where is the space of real symmetric matrices. The space is partially ordered by the Lwner order, that is, for if and only if is positive semidefinite. The inner product in is defined by , where is the trace operation.
Let . Then is self-dual, that is,
Let , , Then is a linear operator from to and its dual is defined by
for any Clearly, is the feasible set of SDP.
Let be a convex function.
(1)The subdifferential of at where , is given by
where is the scalar product on .
(2)The -subdifferential of at is given by
Let Then is called an -approximate solution of SDP, if, for any ,
The conjugate function of a function is defined by
The epigraph of a function , , is defined by
If is sublinear (i.e., convex and positively homogeneous of degree one), then , for all . If , , , then . It is worth nothing that if is sublinear, then
Moreover, if is sublinear and if , , and , then
Let be a closed convex set in and .
(1)Let . Then is called the normal cone to at .
(2)Let . Let . Then is called the -normal set to at .
(3)When is a closed convex cone in , we denoted by and called the negative dual cone of .
Let be a convex function and let be the indicator function with respect to a closed convex subset C of , that is, if , and if . Let . Then
Proposition 2.7 (see ).
Let be a continuous convex function and let be a proper lower semicontinuous convex function. Then
Following the proof of Lemma in , we can prove the following lemma.
Let . Suppose that Let and . Then the following are equivalent:
3. -Duality Theorem
Now we give -duality theorems for SDP. Using Lemma 2.8, we can obtain the following lemma which is useful in proving our -strong duality theorems for SDP.
Let . Suppose that
is closed. Then is an -approximate solution of SDP if and only if there exists such that for any ,
() Let be an -approximate solution of SDP. Then , for any . Let . Then , for any . Thus we have, from Proposition 2.7,
and hence, . So there exists such that and hence there exists such that for any . Since , for any ; and hence it follows from Lemma 2.8 that
Thus there exist , and such that
for any . Thus we have
for any .
() Suppose that there exists such that
for any . Then we have
for any Thus , for any . Hence is an -approximate solution of SDP.
Now we formulate the dual problem SDD of SDP as follows:
We prove -weak and -strong duality theorems which hold between SDP and SDD.
Theorem 3.2 (-weak duality).
For any feasible solution of SDP and any feasible solution of SDD,
Let and be feasible solutions of SDP and SDD respectively. Then and there exist and such that . Thus, we have
Theorem 3.3 (-strong duality).
is closed. If is an -approximate solution of SDP, then there exists such that is a -approximate solution of SDD.
Let be an -approximate solution of SDP. Then for any By Lemma 3.1, there exists such that
for any . Letting in (3.14), . Since and , .
Thus from (3.14),
for any . Hence is an -approximate solution of the following problem:
and so, , and hence, by Proposition 2.6, there exist , such that and
So, is a feasible solution of SDD. For any feasible solution of SDD,
Thus is a 2-approximate solution to SDD.
Now we characterize the -normal set to .
Let and Then
Let and . Then
Let (where is at the th position in )
Thus, we have
From Proposition 3.4, we can calculate .
Let and Then following hold.
(i)If , then
(ii)If and , then
(iii)If and , then
(iv)If and and , then
Now we give an example illustrating our -duality theorems.
Consider the following convex semidefinite program.
and 0. Let and
Then is the set of all feasible solutions of SDP and the set of all -approximate solutions of SDP is . Let . Then is the set of all feasible solution of SDD. Now we calculate the set .
Thus . We can check that for any and any ,
that is, -weak duality holds.
Let be an -approximate solution of SDP. Then and . So, we can easily check that .
Since , from (3.29),
for any . So is an -approximate solution of SDD. Hence -strong duality holds.
Jeyakumar V, Dinh N: Avoiding duality gaps in convex semidefinite programming without Slater's condition. In Applied Mathematics Report. University of New South Wales, Sydney, Australia; 2004.
Ramana MV, Tunçel L, Wolkowicz H: Strong duality for semidefinite programming. SIAM Journal on Optimization 1997, 7(3):641–662. 10.1137/S1052623495288350
Jeyakumar V, Lee GM, Dinh N: New sequential Lagrange multiplier conditions characterizing optimality without constraint qualification for convex programs. SIAM Journal on Optimization 2003, 14(2):534–547. 10.1137/S1052623402417699
Jeyakumar V, Nealon MJ: Complete dual characterizations of optimality for convex semidefinite programming. In Constructive, Experimental, and Nonlinear Analysis (Limoges, 1999), CMS Conference Proceedings. Volume 27. American Mathematical Society, Providence, RI, USA; 2000:165–173.
Dinh N, Jeyakumar V, Lee GM: Sequential Lagrangian conditions for convex programs with applications to semidefinite programming. Journal of Optimization Theory and Applications 2005, 125(1):85–112. 10.1007/s10957-004-1712-8
Jeyakumar V, Lee GM, Dinh N: Lagrange multiplier conditions characterizing the optimal solution sets of cone-constrained convex programs. Journal of Optimization Theory and Applications 2004, 123(1):83–103.
Jeyakumar V, Lee GM, Dinh N: Characterizations of solution sets of convex vector minimization problems. European Journal of Operational Research 2006, 174(3):1380–1395. 10.1016/j.ejor.2005.05.007
Govil MG, Mehra A: -optimality for multiobjective programming on a Banach space. European Journal of Operational Research 2004, 157(1):106–112. 10.1016/S0377-2217(03)00206-6
Gutiérrez C, Jiménez B, Novo V: Multiplier rules and saddle-point theorems for Helbig's approximate solutions in convex Pareto problems. Journal of Global Optimization 2005, 32(3):367–383. 10.1007/s10898-004-5904-4
Hamel A: An -lagrange multiplier rule for a mathematical programming problem on Banach spaces. Optimization 2001, 49(1–2):137–149. 10.1080/02331930108844524
Jeyakumar V, Glover BM: Characterizing global optimality for DC optimization problems under convex inequality constraints. Journal of Global Optimization 1996, 8(2):171–187. 10.1007/BF00138691
Kim GS, Lee GM: On -approximate solutions for convex semidefinite optimization problems. Taiwanese Journal of Mathematics 2007, 11(3):765–784.
Liu JC: -duality theorem of nondifferentiable nonconvex multiobjective programming. Journal of Optimization Theory and Applications 1991, 69(1):153–167. 10.1007/BF00940466
Liu JC: -Pareto optimality for nondifferentiable multiobjective programming via penalty function. Journal of Mathematical Analysis and Applications 1996, 198(1):248–261. 10.1006/jmaa.1996.0080
Strodiot J-J, Nguyen VH, Heukemes N: -optimal solutions in nondifferentiable convex programming and some related questions. Mathematical Programming 1983, 25(3):307–328. 10.1007/BF02594782
Yokoyama K, Shiraishi S: An -optimal condition for convex programming problems without Slater's constraint qualifications. preprint
Hiriart-Urruty JB, Lemarechal C: Convex Analysis and Minimization Algorithms. I. Fundamentals, Grundlehren der mathematischen Wissenschaften. Volume 305. Springer, Berlin, Germany; 1993.
Hiriart-Urruty JB, Lemarechal C: Convex Analysis and Minimization Algorithms. II. Advanced Theory and Bundle Methods, Grundlehren der mathematischen Wissenschaften. Volume 306. Springer, Berlin, Germany; 1993.
This work was supported by the Korea Science and Engineering Foundation (KOSEF) NRL Program grant funded by the Korean government (MEST)(no. R0A-2008-000-20010-0).
About this article
Cite this article
Lee, G., Lee, J. -Duality Theorems for Convex Semidefinite Optimization Problems with Conic Constraints. J Inequal Appl 2010, 363012 (2010). https://doi.org/10.1155/2010/363012
- Approximate Solution
- Feasible Solution
- Convex Function
- Linear Matrix Inequality
- Constraint Qualification