In this paper we consider a Bayesian nonparametric approach to the analysis of discrete-time queueing models. The main motivation consists in applications to telecommunications, and in particular to asynchronous transfer mode (ATM) systems. Attention is focused on the posterior distribution of the overflow rate. Since the exact distribution of such a quantity is not available in a closed form, an approximation based on "proper" Bayesian bootstrap is proposed, and its properties are studied. Some possible alternatives to proper Bayesian bootstrap are also discussed. Finally, an application to real data is provided. (C) 2002 Elsevier B.V. All rights reserved.
Bootstrap approximations for Bayesian analysis of Geo/G/1 discrete-time queueing models / Conti, Pier Luigi. - In: JOURNAL OF STATISTICAL PLANNING AND INFERENCE. - ISSN 0378-3758. - STAMPA. - 120:1-2(2004), pp. 65-84. [10.1016/s0378-3758(02)00499-8]
Bootstrap approximations for Bayesian analysis of Geo/G/1 discrete-time queueing models
CONTI, Pier Luigi
2004
Abstract
In this paper we consider a Bayesian nonparametric approach to the analysis of discrete-time queueing models. The main motivation consists in applications to telecommunications, and in particular to asynchronous transfer mode (ATM) systems. Attention is focused on the posterior distribution of the overflow rate. Since the exact distribution of such a quantity is not available in a closed form, an approximation based on "proper" Bayesian bootstrap is proposed, and its properties are studied. Some possible alternatives to proper Bayesian bootstrap are also discussed. Finally, an application to real data is provided. (C) 2002 Elsevier B.V. All rights reserved.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
