Parallélisation d'une combinaison des méthodes de Monte-Carlo et quasi-Monte-Carlo et application aux réseaux de files d'attente
Parallélisation d'une Combinaison des Méthodes de Monte-Carlo et Quasi-Monte-Carlo et Application aux Réseaux de Files d'Attente
We propose a parallel algorithm which uses both Monte-Carlo and quasi-Monte-Carlo methods. A detailed analysis of this algorithm, followed by examples, shows that the estimator's efficiency is a linear function of the processor number. As a concrete application example, we evaluate performance measures of a multi-class queueing network in steady state.
Particle swarm optimization with various inertia weight variants for optimal power flow solution.
Path functionals over Wasserstein spaces
Given a metric space we consider a general class of functionals which measure the cost of a path in joining two given points and , providing abstract existence results for optimal paths. The results are then applied to the case when is aWasserstein space of probabilities on a given set and the cost of a path depends on the value of classical functionals over measures. Conditions for linking arbitrary extremal measures and by means of finite cost paths are given.
Polynomial-time computability of the edge-reliability of graphs using Gilbert's formula.
Problème de la bipartition minimale d'un graphe