Redundancies in positive-semidefinite quadratic programming

The question investigated is how to detect nonactive restrictions in positive-semidefinite quadratic programming. If the optimization satisfies some regulatory conditions, we can use parametric optimization techniques for that analysis. It turns out that results obtained in an earlier paper, where only positive-definite matrices are considered, can be generalised to the semidefinite case. Simple calculations based exclusively on the problem data allow one to delete superfluous restrictions for this problem class during an optimization procedure.