Colloquium on March 19, 2012Roger Guimera Statistical inference for the discovery of hidden interactions in complex networks In complex systems, individual components interact with each other giving rise to complex networks, which are neither totally regular nor totally random. Because of the interplay between network topology and dynamics, it is crucial to characterize the structure of complex networks. Although during the last decade significant progress has been made in the study of complex networks, we are still far from the ultimate goals of understanding the precise mechanisms responsible for the observed topology, and evaluating the impact of the structure of the network on the dynamics of the system. The two main impairments to achieve these goals are: (i) most network data are very unreliable, that is, for most systems there is uncertainty as to what is the real structure of the network; and (ii) we lack the tools to extract the relevant information contained in the structure of networks, and to evaluate the impact of network structure on a system's dynamics. In my talk, I will discuss how we can use very general properties of complex networks to address these two very prominent problems, and even to go one step further and uncover previously unknown interactions. This opens the door to radically new applications of network theory, including, for example, the prediction of human decisions and preferences, and of novel drug interactions. 


