Statistical Search for The Global Extremum in The Criterion Space

The problem of the global extremum search for non-smooth, multi-extremal objective functions on a bounded set in the R pace still faces a number of problems, especially in multidimensional problems. In this paper we propose a new method for a global extremum search, which is the development of the known psi-transformation method. The method is based on the reduction of the extremum finding process in the criterion space and does not require the introduction of any special assumptions, in particular, information about the Lipschitz constant is not required.