Skip to main content


Tuesday 02 Feb 2021Network Inference Based on Mutual Information Rate

Chris Antonopoulos - University of Essex Meeting ID: 910 8995 5296 Password: 183833 13:30-14:30

We understand a complex system as a system with a large number of interacting components who’s aggregate behaviour is non-linear and undetermined from the behaviour of the individual components. If we now consider these components as nodes in a network, and the underlying physical interaction between any two nodes as links, one way to understand these complex systems is by studying its topological structure, namely, the network connectivity. In this talk, I will present a network inference method based on recorded data and information theory tools, namely the mutual information rate that originates from Shannon entropy. Using recorded data from the nodes in the network, the method can identify the undirected links among them, revealing the underlying network structure. Applications range from functional network inference in neuroscience to financial market networks to social media networks to any other networks based on recorded data sets, to name a few

Add to calendar

Add to calendar (.ics)