Abstract:
The article presents a model of approximation of multidimensional functions based on the apparatus of
multiagent existential mappings (MEM). The applicability theorems of MEM in problems of approximation
of functions of several variables are formulated and proved. The developed model is a constructive combination of the functionality of bioinspired methods and has the following advantages: universality, autonomy of the approximation process, as well as the ability to control the accuracy and time of the learning
process