School of Informatics, Computing and Cyber Systems
Northern Arizona University
Process Networks are a simple and intuitive concept with powerful applications in science. A dynamical process network is a weighted directed graph where nodes represent observable subsystems and links represent processes of change. Dynamical process networks can easily be delineated empirically for any observed system of model of a system, and the geometry of the network reveals the system's function at the granularity of individual processes. When a process network is constructed using information flow, the resulting process links can be interpreted as cause-effect connections that also imply statistical predictability in the Bayesian sense. Circular connections indicate feedback and synchronization, a key property of complex and self-organizing systems. Process networks have been historically applied to dynamical systems modeling, neuroscience, observatory science, and ecohydrology problems, and are now beginning to be applied for the purpose of hypothesis testing and scientific model critique. This talk presents a brief history of process networks, their applications, and tools for scientists."