  


On the identification of optimal partition for semidefinite optimization
Ali MohammadNezhad (alm413lehigh.edu) Abstract: The concept of the optimal partition was originally introduced for linear optimization and linear complementarity problems and subsequently extended to semidefinite optimization. For linear optimization and sufficient linear complementarity problems, the optimal partition and a maximally complementary optimal solution can be identified in strongly polynomial time. In this paper, under no assumption on strict complementarity, we consider the identification of the optimal partition of semidefinite optimization for solutions on, or in a neighborhood of the central path. In contrast to linear optimization and linear complementarity problem, the optimal partition for semidefinite optimization cannot be identified exactly from an interior solution. Instead, we identify the sets of eigenvectors converging to an orthonormal bases for the optimal partition using the bounds for the magnitude of the eigenvalues. The magnitude of the eigenvalues of an interior solution is quantified using a condition number and an upper bound for the distance of an interior solution to the optimal set. We provide iteration complexity bound for the identification of the above sets of eigenvectors. Keywords: Conic programming and interior point methods, Semidefinite optimization, Optimal partition, Maximally complementary optimal solution, Degree of singularity Category 1: Linear, Cone and Semidefinite Programming (Semidefinite Programming ) Category 2: Linear, Cone and Semidefinite Programming Category 3: Convex and Nonsmooth Optimization (Convex Optimization ) Citation: Report No. 17T008, Industrial and Systems Engineering Department, Lehigh University, June 2017. Download: [PDF] Entry Submitted: 09/09/2016 Modify/Update this entry  
Visitors  Authors  More about us  Links  
Subscribe, Unsubscribe Digest Archive Search, Browse the Repository

Submit Update Policies 
Coordinator's Board Classification Scheme Credits Give us feedback 
Optimization Journals, Sites, Societies  