Show simple item record

dc.contributor.authorColeman, Thomas F.en_US
dc.contributor.authorHulbert, Laurieen_US
dc.description.abstractWe present a globally and superlinearly convergent algorithm for solving convex quadratic programs with simple bounds. We develop our algorithm using a new formulation of the problem: the minimization of an unconstrained piecewise quadratic function that has the same optimality conditions as the original problem. The major work at each iteration is the Cholesky factorization of a positive definite matrix with the size and structure of the Hessian of the quadratic. Hence our algorithm is suitable for solving large sparse problems and for implementation on parallel computers. We implemented our algorithm and tested it on a sequential computer on a variety of dense problems, and we present numerical results which show that our algorithm solves many problems quickly. Keywords: quadratic programming, interior point methods, simple bounds, box constraints, large sparse minimization.en_US
dc.format.extent2312235 bytes
dc.format.extent589561 bytes
dc.publisherCornell Universityen_US
dc.subjectcomputer scienceen_US
dc.subjecttechnical reporten_US
dc.titleA Globally and Superlinearly Convergent Algorithm for Convex Quadratic Programs with Simple Boundsen_US
dc.typetechnical reporten_US

Files in this item


This item appears in the following Collection(s)

Show simple item record