Particle Swarm Optimization
Also known as PSO
Overview
Particle Swarm Optimization studies how a population of candidate solutions can search for good answers by updating each candidate using its own experience and information shared by other particles. Common approaches change the swarm’s interaction topology or diversity, combine PSO with gradients or local search, and use surrogate models or learned guidance to reduce costly evaluations. Researchers apply these methods to continuous and discrete benchmarks, engineering design, neural architecture search, and tracking changing optima. A central challenge is balancing exploration with convergence: methods that help on one landscape may fail on another, making robust performance and fair comparisons under equal evaluation budgets ongoing concerns.
Latest papers
21 papers2 papers in the last four weeks, against 1 the four weeks before. 0.0% of all new papers.