A Numerical Comparison of Consensus‐Based Global Optimization to other Particle‐based Global Optimization Schemes
Claudia Totzeck, René Pinnau, Sebastian Blauth, Steffen Schotthöfer · PAMM · 2018
Abstract We compare a first‐order stochastic swarm intelligence model called consensus‐based optimization (CBO), which may be used for the global optimization of a function in multiple dimensions, to other particle swarm algorithms for global optimization. CBO allows for passage to the mean‐field limit resulting in a nonlocal, degenerate, parabolic PDE. Exploiting tools from PDE analysis, it is possible to rigorously prove convergence results for the algorithm (see [3]). In the present article we discuss numerical results obtained with the Particle Swarm Optimization (PSO) [4], Wind‐Driven Optimization (WDO) [6] and CBO and show that CBO leads to very competitive results.