Particle Swarm Optimization Investigation of Social Interaction Between Agent Particles by: Kenneth Lee
An Overview of PSO History Implementation Processes Expected Results My Project An Overview of PSO History Implementation Processes Expected Results My Project
History of PSO Model of Flocks of Birds Early Inefficiencies Implementation Process Expected Results My Project History of PSO Model of Flocks of Birds Early Inefficiencies Velocity Matching Craziness Refinement http://www.richard-seaman.com
Implementations of PSO History Implementation Process Expected Results My Project Implementations of PSO Used to find Function Minimums N-Dimensional Infinite search spaces Neural Networking
Process of PSO Initialize Swarm Move Swarm Calculate pbest and gbest History Implementation Process Expected Results My Project Process of PSO Initialize Swarm Move Swarm Calculate pbest and gbest Adjust Velocities Cognitive Social Convergence
Expected Results of PSO History Implementation Process Expected Results My Project Expected Results of PSO Convergence Correctness Cheap Memory Time
My Project Social Interactions How do the particle talk to each other? History Implementation Process Expected Results My Project My Project Social Interactions How do the particle talk to each other? Can Different ways of talking improve performance? Cost?
My Project – Current State History Implementation Process Expected Results My Project My Project – Current State Rastrigin Cornfield Six Camel Hump Current Benchmark Functions
My Project – Problems FIPS (Fully Informed Particle Swarm) History Implementation Process Expected Results My Project My Project – Problems FIPS (Fully Informed Particle Swarm) Converges on (0,0) instead of the minimum values Not enough social interactions or benchmark testing functions