Approximate Bayesian computation, an introduction
Résumé
Approximate Bayesian Computation (ABC) methods have become a “mainstream” statistical technique in the past decade, following the realisation that they were a form of non-parametric inference, and connected as well with the econometric technique of indirect inference. In this survey of ABC methods, we focus on the basics of ABC and cover some of the recent literature, following our earlier survey in Marin et al.(2011). Given the recent paradigm shift in the perception and practice of ABC model choice, we insist on this aspect of ABC techniques, including in addition some convergence results.