Admission control in a pure loss healthcare network: MDP and DES approach - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2013

Admission control in a pure loss healthcare network: MDP and DES approach

Résumé

(This paper considers admission control policies in a pure-loss hierarchical perinatal network where there are 2 parallel multi-server (target) hospitals fed by new arriving patients and overflowed patients from a set of parallel multi-server hospitals. In this perinatal network setting, we consider the problem of finding an optimal admission policy that recommends how many beds to reserve in two target hospitals for each arriving stream in order to maximize total revenue in the system. At first, we assumed a Markovian system and model the system as a Markov Decision Process (MDP). By using value iteration algorithm, optimal admission policy is computed. Afterwards, we evaluate various policy scenarios (including MDP optimal policy) with a simulation model which strengthens the decision making process by incorporating the complexity which cannot be captured by MDP and we assess the impact of Markovian assumption in a complex healthcare setting.)
Fichier non déposé

Dates et versions

hal-00923993 , version 1 (06-01-2014)

Identifiants

  • HAL Id : hal-00923993 , version 1

Citer

Canan Pehlivan, Vincent Augusto, Xiaolan Xie. Admission control in a pure loss healthcare network: MDP and DES approach. WSC 2013 (Winter Simulation Conference), Dec 2013, Washington, DC, United States. ⟨hal-00923993⟩
197 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More