In this chapter we discuss applications of machine learning (ML) to the field of quantum-mechanical simulations. Since the discovery of the quantum mechanics more than hundred years ago, a large number of computational methods have been developed to enable the calculation the physical and chemical properties of molecules and solids. One of the most used first-principles methods is the density functional theory (DFT). Compared to similar methods, it provides a reasonable approximation while keeping manageable computational costs. Nowadays, it can even be used as a black box tool which allows for high-throughput studies. The abundance of data makes DFT appealing for machine learning. On the one hand, it enables the learning of general structure-property relations with or without the aid of human theories, and, on the other hand, these data can be used to develop ML models for the functionals employed in the quantum-mechanical calculations.