Precision in High Dimensional Optimisation of Global Tasks with Unknown Solutions

Abstract

High dimensional optimisation is a challenge for most of the available search methods. Resolving global and constrained task seems to be even harder and exploration of tasks with unknown solutions can be seen very rare in the literature and requires more research efforts. This article analyses optimisation of high dimensional global, including constrained, tasks with unknown solutions. Reviewed and analysed are experimental results precision, possibilities for trapping in local sub-optima and adaptation to unknown search spaces.