1: \begin{abstract}

2:     We propose and analyze the convergence of a novel stochastic algorithm for solving monotone inclusions that are the sum of a maximal monotone operator and a monotone, Lipschitzian operator. The propose algorithm requires only unbiased estimations of the Lipschitzian operator.

3:      We obtain the rate $\mathcal{O}(log(n)/n)$ in expectation for the strongly monotone case, as well as almost sure convergence for the general case. 

4:      Furthermore, in the context of  application to convex-concave saddle point problems, we derive the rate of the primal-dual gap. In particular,

5:      we also obtain $\mathcal{O}(1/n)$ rate convergence of the primal-dual gap in the deterministic setting.

6: \end{abstract}

7: