Controllability robustness of complex networks

This article presents an overview on the state-of-the-art development in complex network controllability and its robustness against malicious attacks and random failures. Specifically, it first reviews the concepts of network pinning control and controllability, and then discusses the network controllability robustness against destructive attacks by means of node- and/or edge-removal. The related issue of network connectivity robustness is also discussed. To that end, it furthermore provides an brief overview on the recent development of a machine-learning approach for predicting optimal network controllability robustness, which may shed some lights on the understanding of optimal network structures for various design considerations.