1: \begin{abstract}

2: \singlespace

3: Approximate Bayesian Computation (ABC) can be viewed as an analytic

4: approximation of an intractable likelihood coupled with an elementary

5: simulation step. \amj{Such a view, combined with a suitable instrumental prior

6:   distribution permits} maximum-likelihood (or

7: maximum-a-posteriori) inference to be conducted, approximately, using

8: essentially the same techniques. \amj{An elementary} approach \amj{to this

9:   problem which simply obtains a nonparametric approximation of the likelihood

10:   surface which is then used as a smooth proxy for the likelihood in a

11:   subsequent maximisation step} is developed here and the convergence of this

12: class of algorithms is characterised theoretically. The use of non-sufficient

13: summary statistics {in this context} is considered. Applying

14: the proposed method to {four} problems demonstrates good

15: performance. The proposed approach provides an alternative for approximating

16: the maximum likelihood estimator (MLE) in complex scenarios.

17: 

18: \medskip

19: 

20: \noindent\textbf{Keywords: } Approximate Bayesian Computation; Density Estimation;

21: Maximum Likelihood Estimation; Monte Carlo Methods.

22: \end{abstract}