Optimization Online


Interior Point Methods for Computing Optimal Designs

Zhaosong Lu (zhaosong***at***sfu.ca)
Ting Kei Pong (tkpong***at***math.washington.edu)

Abstract: In this paper we study interior point (IP) methods for solving optimal design problems. In particular, we propose a primal IP method for solving the problems with general convex optimality criteria and establish its global convergence. In addition, we reformulate the problems with A-, D- and E-criterion into linear or log-determinant semidefinite programs (SDPs) and apply standard primal-dual IP solvers such as SDPT3 [21,25] to solve the resulting SDPs. We also compare the IP methods with the widely used multiplicative algorithm introduced by Silvey et al. [18]. The computational results show that the IP methods generally outperform the multiplicative algorithm both in speed and solution quality. Moreover, our primal IP method theoretically converges for general convex optimal design problems while the multiplicative algorithm is only known to converge under some assumptions.

Keywords: Optimal design, A-criterion, c-criterion, D-criterion, E-criterion, pth mean criterion, interior point methods

Category 1: Convex and Nonsmooth Optimization (Convex Optimization )

Category 2: Linear, Cone and Semidefinite Programming (Semi-definite Programming )

Category 3: Applications -- Science and Engineering (Statistics )

Citation: Manuscript, Department of Mathematics, Simon Fraser University, 8888 University Drive, Burnaby, BC, V5A 1S6, Canada, September 2010

Download: [PDF]

Entry Submitted: 09/07/2010
Entry Accepted: 09/07/2010
Entry Last Modified: 03/27/2011

Modify/Update this entry

  Visitors Authors More about us Links
  Subscribe, Unsubscribe
Digest Archive
Search, Browse the Repository


Coordinator's Board
Classification Scheme
Give us feedback
Optimization Journals, Sites, Societies
Mathematical Programming Society