Variable Metric Relaxation Methods. Part I. A Conceptual Algorithm.
Abstract
A variable metric, maximal ellipsoidal distance, relaxation method for solving systems of linear, inequalities is described and its rate of convergence analyzed. This algorithm, with a termination routine appended to it, is polynomial if the ellipsoid matrix approximates an inverse 'Hessian,' which may be defined through the largest ellipsoid contained in a perturbation cum compactification of the feasible set. This so called inverse Hessian is given by a positive linear combination of symmetric rank one matrices built on the vectors given by the facets of the perturbation cum compactification of the feasible set. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1981
- Accession Number
- ADA107461
Entities
People
- Jean-louis Goffin
Organizations
- Stanford University