Seminarium "Teoria cząstek elementarnych i kosmologia"

Nowadays, “dataism”, i.e. the philosophy created by the emergence of the significance of the Big Data, starts to dominate contemporary research and high-technology applications. The crucial question arises, how, in such gargantuan sets of data, to unravel signals from noise and to find the hidden, usually non-linear, correlations. Random matrix theory, initially used for description of excited heavy nuclei, and later, successfully applied in all domains of physics, ranging from disorder in quantum systems to quantumchromodynamics and quantum gravity, seems to be the right tool. In this lecture I argue, that in the limit of large dimension of the matrix (analogue of ‘t Hooft limit in QFT), such theory can be viewed as a XXI century counterpart of the XIX century classical probability theory. I demonstrate sample interdisciplinary applications of such a viewpoint in various real complex systems.