this is Particle_Swarm_Optimization_PPT.pptx

This is particle swarm optimization

1. Particle Swarm Optimization (PSO) Nature-Inspired Metaheuristic Algorithm Your Name | Course Info

2. Introduction • Developed by Kennedy & Eberhart (1995) • Inspired by bird flocking & fish schooling • Belongs to Swarm Intelligence family • Useful for continuous & combinatorial optimization

3. Biological Inspiration • Particles represent birds/fish searching for food • Each adjusts position based on: - Own best experience (pBest) - Swarm’s best experience (gBest) • Collective learning → optimization

4. PSO Algorithm Steps 1. Initialize particles randomly (position & velocity) 2. Evaluate fitness function 3. Update personal best (pBest) and global best (gBest) 4. Update velocity & position 5. Repeat until convergence

5. Equations Velocity Update: vi(t+1) = w*vi(t) + c1*r1*(pBesti - xi(t)) + c2*r2*(gBest - xi(t)) Position Update: xi(t+1) = xi(t) + vi(t+1) w = inertia weight c1, c2 = cognitive & social parameters r1, r2 = random factors

6. Key Concepts • pBest = best position a particle has visited • gBest = best position found by entire swarm • Exploration vs Exploitation balance

7. Applications • Function optimization • Neural network training • Feature selection in ML • Job scheduling • Robotics path planning • Image processing

8. Advantages • Easy to implement • Few parameters to tune • Works well for nonlinear problems

9. Limitations • Can get stuck in local optima • Sensitive to parameter tuning • Less effective in very high-dimensional problems

10. Variants of PSO • Global Best (gBest) model • Local Best (lBest) model • Constriction Coefficient PSO • Hybrid PSO with GA, ACO, DE

11. Case Study Example • Example: Optimize Rosenbrock or Sphere function • Show convergence of swarm towards optimum

12. Conclusion • PSO = powerful swarm intelligence algorithm • Efficient for global optimization • Future: Hybrid & Adaptive PSO approaches

13. References • Kennedy, J., & Eberhart, R. (1995). Particle Swarm Optimization. IEEE Proc. • Clerc, M., & Kennedy, J. (2002). The Particle Swarm – Explosion, Stability, and Convergence.