Today, neural networks are actively used in modeling complex nonlinear dependencies. Amid such a rapid growth of interest in this powerful tool for modeling various objects and processes, research in the natural sciences and engineering, the work on the application of neural networks in economics is vanishingly small. This is explained both by the complexity of the modeling tool itself - neural networks, and by the object of modeling - the evolving economy. At the dawn of the development of neural networks, the method of modeling processes using Kolmogorov-Gabor polynomials (or Wiener series) was considered as an alternative. For various reasons, this method lost the competitive battle, and neural networks prevail. The article presents a method and technique for constructing an elementary image of the Kolmogorov-Gabor polynomial and shows that today it can be used as an alternative to neural networks in modeling of economic processes.