<p>Today, determining mechanical behavior of polycrystalline materials is possible through experimental studies; however, the approach requires considerable financial resources and specific equipment. The other way to handle the above problem is through mathematical modeling and data analysis. This paper discusses the method of creating datasets that describe the mechanical behavior of two-component polycrystalline materials. Also, the data analysis is provided, which allows for determining the probabilistic characteristics of the elastic constants as well as the yield surface. To build the datasets, mathematical modeling capabilities, in particular the DAMASK virtual lab, are used. The microstructure of the material is constructed using the probabilistic cellular automata method. The datasets obtained are analyzed using mathematical statistics. The probability density functions as well as their parameters are determined. Based on the statistical data, the probabilistic yield surface is constructed and the probability of occurrence of plastic strains is determined. As a result of the work, the values of probabilistic characteristics of elastic constants and the yield surface are obtained. Also, the probability of occurrence of the plastic strains is determined. To demonstrate the method, the stress analysis is performed using a plate with a hole.</p>