On the applicability and limitations of formal verication of machine learning systems

The paper deals with the issues of formal verication of machine learning systems. With the growth of the introduction of systems based on machine learning in the so-called critical systems (systems with a very high cost of erroneous decisions and actions), the demand for conrmation of the stability of such systems is growing. How will the built machine learning system perform on data that is dierent from the set on which it was trained? Is it possible to somehow verify or even prove that the behavior of the system, which was demonstrated on the initial dataset, will always remain so? There are dierent ways to try to do this. This article deals with the formal verication of machine learning systems.