Accurate solution of dense linear systems, Part II: Algorithms using directed rounding
Journal of Computational and Applied Mathematics
Hi-index | 0.00 |
We prove that the set of optimal basic variables of a linear program remains stable under mutually independent variations of all data within prescribed tolerances if and only if it is stable for a finite subset of explicitly described linear programs from this family. The cardinality of this subset is exponential in the number of constraints.