A new parameterized logarithmic kernel function for linear optimization with a double barrier term yielding the best known iteration bound
Résumé
In this paper, we propose a large-update primal-dual interior point algorithm for linear optimization. The method is based on a new class of kernel functions which differs from the existing kernel functions in which it has a double barrier term. The investigation according to it yields the best known iteration bound O
n
log
(
n
)
log
(
n
∈
)
\sqrt n \log (n)\log \left( {{n \over \in }} \right)
for large-update algorithm with the special choice of its parameter m and thus improves the iteration bound obtained in Bai et al. [2] for large-update algorithm.
Domaines
Mathématiques [math]| Origine | Accord explicite pour ce dépôt |
|---|
