1: \begin{abstract}

2:   We study a perturbed version of the proximal gradient algorithm for which the

3:   gradient is not known in closed form and should be approximated. We address the convergence and derive a

4:   non-asymptotic bound on the convergence rate for the perturbed proximal

5:   gradient, a perturbed averaged version of the proximal gradient algorithm and

6:   a perturbed version of the fast iterative shrinkage-thresholding (FISTA) of

7:   \cite{becketteboulle09}.  When the approximation is achieved by using Monte Carlo methods,

8:   we derive conditions involving the Monte Carlo batch-size and the

9:   step-size of the algorithm under which convergence is guaranteed.  In

10:   particular, we show that the Monte Carlo approximations of some averaged proximal

11:   gradient algorithms and a Monte Carlo approximation of FISTA  achieve the same

12:   convergence rates as their deterministic counterparts.  To illustrate, we apply

13:   the algorithms to high-dimensional generalized linear mixed models using $\ell_1$-penalization.

14: \end{abstract}

15: