Toward Structural Controllability and Predictability in Directed Networks

The lack of studying the complex organization directed network usually limits understanding underlying relationship between structures and functions. Structural controllability structural predictability, two seemingly unrelated subjects, are revealed in this article to be both highly dependent on critical links previously thought only able influence number driver nodes controllable networks. Here, we show that can not contribute but also have a significant impact predictability networks, suggesting universal pattern reciprocity In addition, it is shown fraction location strong performance prediction algorithms. Moreover, these empirical results interpreted by introducing link centrality based corresponding line graphs. This work bridges gap independent research fields, provides indications developing advanced control strategies algorithms from microscopic perspective.