ABSTRACT
This chapter aims at supplying information about the theoretical basis of Approximate Bayesian Computation (ABC), which is an efficient computational tool to solve inverse problems without the need to formulate, nor evaluate the likelihood function. By ABC, the posterior PDF can be computed in those cases where the likelihood function is intractable, impossible to formulate, or computationally demanding. Several ABC pseudo-codes are included in this chapter and an example of application is provided. Finally, the ABC-SubSim algorithm, which was initially proposed by Chiachío et al. [SIAM Journal of Scientific Computing, Vol. 36, No. 3, pp. A1339–A1358], is explained within the context of an example of application.
