The influence of the sparsity on topology reconstruction.

<p>The estimation error changes with the sparsity of directed random networks. There, the sparsity is defined as the ratio of the number of zero non-diagonal elements to , and is calculated using the -norm convex optimization strategy (with ). Furthermore, each black point represents the result of averaging over 30 random perturbations (with and being uniformly distributed in the range ) and the standard square error is given as well.