Detailed view

Preprint, Working Paper, ...

Attached file list to this document:

PDF abc10.pdf(344.5 KB) PS abc10.ps(999.6 KB)

New Insights into Approximate Bayesian Computation

Gérard Biau1, 2, 3, 4, Frédéric Cérou5, Arnaud Guyader5

Approximate Bayesian Computation (ABC for short) is a family of computational techniques which offer an almost automated solution in situations where evaluation of the posterior likelihood is computationally prohibitive, or whenever suitable likelihoods are not available. In the present paper, we analyze the procedure from the point of view of k-nearest neighbor theory and explore the statistical properties of its outputs. We discuss in particular some asymptotic features of the genuine conditional density estimate associated with ABC, which is a new interesting hybrid between a k-nearest neighbor and a kernel method.

1:	LPMA - Laboratoire de Probabilités et Modèles Aléatoires
2:	LSTA - Laboratoire de Statistique Théorique et Appliquée
3:	DMA - Département de Mathématiques et Applications
4:	INRIA Paris - Rocquencourt - CLASSIC
5:	INRIA - IRISA - ASPI

Research team: Processus stochastique

Keyword(s) : Approximate Bayesian Computation – Nonparametric estimation – Conditional density estimation – Nearest neighbor methods – Mathematical statistics

hal-00721164, version 1

http://hal.archives-ouvertes.fr/hal-00721164

oai:hal.archives-ouvertes.fr:hal-00721164

From: Gérard Biau <>

Submitted on: Thursday, 26 July 2012 16:58:13

Updated on: Monday, 10 September 2012 09:44:25