Robust Critical-like Dynamics in Complex Networks: Facilitation, Path Dependence, and Implications
Abstract
Approved for Public ReleaseEmpirical evidence suggests that biological and technological networks may have evolved to operate near t,he tipping point between quiescence and large-scale perturbation amplification in order to maximize functionality. In particular, th,e ability of the network to respond to stimuli across a range of sizes, i.e., its dynamic range, is expected to be optimal near this, tipping/critical point. However, proximity to criticality as an explanation for observed power law distributions in a wide range of, systems remains controversial, with many critics claiming that precise external- or self- tuning to the critical point is implausib,le. In this project, we explore the dynamical and structural origins of robust critical-like dynamics. We say that a system exhibi,ts robust critical-like dynamics if, for a wide range of parameters away from the phase transition, it displays statistics that appr,oximately match the expected power law behavior at criticality. We hypothesize that systems exhibiting robust critical-like dynamics, are able to leverage the functional advantages of criticality without the need for complex tuning mechanisms. We focus primarily on, the application of these ideas to information processing in real and artificial neural networks, but we believe that these concepts, may also apply to other technological systems like supply chain networks and the Internet of things.The project addresses three key, issues:1.What system properties, in terms of network structures and dynamical interaction rules, serve to facilitate robust critica,l-like dynamics? What advantages does this robustness confer?2.How does the path toward/around criticality (e.g., through evolution,or plasticity) influence the level of robustness the system can achieve?3.How can we design more flexible machine learning schemes b,y incorporating robust critical-like dynamics into artificial neural networks?
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 08, 2022
- Source ID
- N000142212656
Entities
People
- Michelle Girvan
Organizations
- Office of Naval Research
- United States Navy
- University of Maryland