1: \begin{abstract}

2: A constrained $\ell_1$ minimization method is proposed for

3: estimating a sparse inverse covariance matrix

4: based on a sample of $n$ iid $p$-variate random variables.

5: The resulting estimator is shown to enjoy a

6: number of desirable properties. In particular, it is shown that the

7: rate of convergence between the estimator and the true $s$-sparse

8: precision matrix under the spectral norm is $s\sqrt{\log p/n}$ when

9: the population distribution has either exponential-type tails or

10: polynomial-type tails. Convergence rates under the elementwise

11: $\ell_{\infty}$ norm and Frobenius norm are also presented.

12: In addition, graphical model selection is considered.

13: The procedure is easily implementable by linear programming.

14: Numerical performance of the estimator

15: is investigated using both simulated and real data. In particular, the

16: procedure is applied to analyze a breast cancer dataset.

17: The procedure performs favorably in comparison to existing methods.

18: \end{abstract}

19: