Abstract by Levi Pike
Social Network Swarms
In nature, insect, bird, and animal groups (often called swarms, flocks, colonies, etc.) are known to show collective intelligence. The organisms behave in a complex manner as a group, and yet are composed of simple rules at the individual level. These groups solve optimization problems efficiently and robustly in unknown environments, like ants searching for food, or honey bees searching for new places to form nests. Many important problems can be modeled as an optimization over an unknown environment. One such class of problems is the ``Best-M-of-N problem'', in which the top M solutions must be selected from N possible solutions. Doing this without the knowledge of the complete environment is a task that is suitable for distributed systems, specifically, swarms. This paper investigates a novel swarm behavior algorithm for solving the Best-M-of-N problem in a network, which is a non-continuous and non-spatial environment.