Category Archives: Optimization

In which the optimizer eats itself, part I

A vanilla PSO algorithm has a handful of tunable parameters which affect the convergence rate. For our purposes, we're going to consider the set \{ \omega, c_1, c_2, n_{\mathrm{particles}} \} to be our set of adjustable parameters. It is certainly true that the bounds on both position and velocity for each particle will affect convergence,

Termination conditions

One of the projects I'm supervising/working on this year is connected with optimization, and in particular, particle swarm based optimization algorithms. One of the things we've been struggling with is termination conditions, and so I thought I'd write a bit about it to clear my head. Any useful algorithm has to terminate at some point.