Link prediction in complex dynamic networks using multiple interdependent time series

Link Prediction is an important problem in complex network analysis. The link prediction problem in dynamic network is a more difficult problem than its static counterpart, as dynamic networks allow links to be created or deleted over a period of time. We model the link prediction problem in dynamic network as a classification problem and use vector auto-regression based time series models and supervised learning approaches to predict the link status at future. The proposed method shows a significant improvement over the state-of-the-art.