1: \begin{abstract}

2: We present a novel accelerated primal-dual (APD) method for solving a

3: class of deterministic and stochastic saddle point problems (SPP). The basic idea of

4: this algorithm is to incorporate a multi-step acceleration scheme into the primal-dual method

5: without smoothing the objective function. For deterministic SPP, the APD method achieves the

6: same optimal rate of convergence as Nesterov's  smoothing technique.  Our stochastic APD method

7: exhibits an optimal rate of convergence for stochastic SPP not only in terms of its dependence on the

8:  number of  the iteration, but also on a variety of problem parameters. To the best

9: of our knowledge, this is the first time that such an optimal algorithm has been developed for stochastic SPP

10: in the literature. Furthermore,  for both deterministic and stochastic SPP,  the developed APD algorithms

11: can deal with the situation when the feasible region is unbounded, as long as a saddle point exists.

12: In the unbounded case, we incorporate the modified termination criterion introduced by Monteiro and

13: Svaiter in solving SPP problem posed as monotone inclusion, and demonstrate that the rate of convergence of

14: the APD method depends on the distance from the initial point to the set of optimal solutions.

15: \end{abstract}