There have been many achievements in the last decade in the development of materials science, chemistry, and physics (experimental and theoretical). However, the most difficult problem—calculating the intrinsic properties of multicomponent compounds starting from the knowledge of their constituent components' properties—still remains unsolved. Calculations or predictions based on only the properties of constituent components (or simply, properties) are called a priori calculations or predictions. These difficulties are due to the solution of mathematical problems arising in the quantum mechanical calculations of multi-electronic systems. As a result, scientists make use of many empirical prediction methods that use existing regularities from a variety of property data. Some of the empirical criteria for the formation of compounds with predefined properties use the rules of Hume-Rothery, Laves, Mathias, Goldschmidt, Villars, and Darken-Gurry.