Analysis of landscape of optimizable functions when making decisions related to data management

The article states the principle of searching for the most efficient data management solutions by metaheuristic algorithms. Metaheuristics is an iterative procedure that uses randomization and self-learning elements, intensification and diversification of searches, adaptive control mechanisms, constructive heuristics,, and local search methods. It is a very promising approachto solving many optimization problems. A pseudo-code program that illustrates the search for solutions by metaheuristic algorithms is proposed. General landscape model of the optimized function is considered. The distance between the solutions is determined. Various measures are used to analyze the landscape: Changing the average distance between the set of uniformly distributed solutions and the set of locally optimal solutions; entropy; amplitude; correlation