1: \begin{abstract} 

2: We consider the problem of jointly estimating the parameters as well

3: as the structure of binary valued Markov Random Fields, in contrast to

4: earlier work that focus on one of the two problems. We formulate the

5: problem as a maximization of $\ell_1$-regularized surrogate likelihood

6: that allows us to find a sparse solution. Our optimization technique

7: efficiently incorporates the cutting-plane algorithm in order to

8: obtain a tighter outer bound on the marginal polytope, which results

9: in improvement of both parameter estimates and approximation to

10: marginals.  On synthetic data, we compare our algorithm on the two

11: estimation tasks to the other existing methods. We analyze the method

12: in the high-dimensional setting, where the number of dimensions $p$ is

13: allowed to grow with the number of observations $n$.  The rate of

14: convergence of the estimate is demonstrated to depend explicitly on

15: the sparsity of the underlying graph.

16: \end{abstract}