Publication
A recent study shows that neural symbolic regression offers a route to automated discovery of governing equations for network dynamics across high-dimensional complex systems. Complex systems are all around us, from gene regulatory networks controlling cellular function, to interacting ecological communities and the spread of diseases in human and animal populations. Describing these systems and understanding how they evolve requires mathemati-cal laws that capture the interactions between the system’s units and the dynamical processes unfolding on them. The increased abundance of observational data combined with unprecedented computational power has given us access to the heterogeneous and high-dimensional nature of complex networked systems. Yet identifying general and interpretable mathematical laws governing them remains a fundamen-tal challenge.
NetSI authors
