1: \begin{abstract}

2: This paper proposes a new method for estimating sparse precision matrices

3: in the high dimensional setting. This procedure applies a novel Sparse

4: Column-wise Inverse Operator (SCIO) to modified sample covariance

5: matrices. We establish the convergence rates of this procedure under

6: various matrix norms. Under the Frobenius norm loss, we prove theoretical

7: guarantees on using cross validation to pick data-driven tunning parameters.

8: Another important advantage of this estimator is its efficient computation

9: for large-scale problems, using a path-following coordinate descent

10: algorithm we provide. Numerical merits of our estimator are also illustrated

11: using simulated and real datasets. In particular, this method is found

12: to perform favorably on analyzing an HIV brain tissue dataset and

13: an ADHD resting fMRI dataset.

14: \end{abstract}

15: