ИСТИНА

This article analyzes the errors and reasons for the failure of projects using machine learning. Technically, according to academic articles, the percentage of failed projects is quite high. The literature gives figures such as 87% of unsuccessful projects. Naturally, under such conditions, the problem of analyzing such errors becomes very relevant. The article, based on many analyzed works, presents summary data on errors and failures of projects using machine learning and analyzes the relationship of these causes with the requirements for the robustness of designed systems. It is shown that most of the reasons are, in fact, the lack of robustness for machine learning systems. The paper also shows the importance of the transition to data-centric systems, presents forecasts for the further development of machine learning models for critical applications.