Surrogate Based Optimization (SBO)
Find the Best Parameter Combination – in the Shortest Time!
Surrogate based optimization enables short turnaround times within your development process. The necessary simulations can be carried out completely in parallel and subsequently reused for a change in the target function.
Based on the surrogate, the expected improvement is evaluated at each iteration in order to select new, promising parameter combinations. Its mathematical characteristics produce a favorable balance between exploration and exploitation and prevent an unnecessarily large number of samples at putative optima.