1: \begin{abstract}

2: Dual averaging-type methods are widely used in industrial machine learning applications due to their ability to promoting solution structure (e.g., sparsity) efficiently. In this paper, we propose a novel accelerated dual-averaging primal-dual algorithm for minimizing a composite convex function.

3: We also derive a stochastic version of the proposed method which solves empirical risk minimization, and its advantages on handling sparse data are demonstrated both theoretically and empirically.

4: %The analysis shows that it has accelerated convergence rates in several different settings. We also study the case where the objective function has a finite-sum structure, and develop a stochastic version for this case, which is also proven to have optimal convergence rate.

5: \end{abstract}