Physics-Inspired Algorithms and Phase Transitions in Community Detection - 2014 Symposium

Physics-Inspired Algorithms and Phase Transitions in Community Detection - 2014 Symposium

Explore a physics-inspired approach to community detection in this 48-minute lecture from the Santa Fe Institute's Annual Science Board Symposium. Delve into the stochastic block model, statistical inference, and belief propagation techniques. Examine phase transitions in community detection, from detectable to undetectable, and investigate hierarchical clustering methods. Learn about the non-backtracking operator and its applications in trust and centrality analysis. Gain insights into the intersection of physics culture and machine learning, and discover the challenges and opportunities in this field.