Explore the fascinating world of swarm intelligence in this 24-minute conference talk from Churchill CompSci Talks. Delve into the principles of nature-inspired computational techniques, focusing on the application of swarm intelligence to solve complex problems. Learn how to develop a simple yet accurate model of bird flocking behavior and discover how this model can be extended to tackle numerical optimization challenges that traditional methods struggle with. Gain insights into the popular 'Particle Swarm Optimisation' algorithm, understanding its mathematical foundations and practical applications. Through a practical demonstration, witness the power of this distributed system in action and grasp key takeaways that highlight the potential of bio-inspired algorithms in computer science.
Overview
Syllabus
Intro
CONTENTS
EMERGENT COMPLEXITY
COMPLEXITY IN ARTIFICIAL SYSTEMS
SWARM INTELLIGENCE
A MINIMAL FLOCKING MODEL
SEPARATION
ALIGNMENT
COHESION
THE OPTIMISATION PROBLEM
PARTICLE SWARM OPTIMISATION
MATHEMATICAL FORM
THE PSO ALGORITHM
PSO AS A DISTRIBUTED SYSTEM
PRACTICAL DEMONSTRATION
TAKE-AWAY POINTS
Taught by
Churchill CompSci Talks
