Deep learning complete intersection Calabi-Yau manifolds

In recent years, deep learning has become a relevant research theme in physics and mathematics. It is a very efficient method for data processing, and elaboration and exploration of patterns [1]. Though the basic building blocks are not new [2], the increase in computational capabilities and the creation of larger databases lead new deep learning techniques to thrive. Specifically, the understanding of the geometrical structures [3, 4] and the representation learning [5] are of particular interest from a mathematical and theoretical physics points of view [6-9]. We are interested in applications of data science and deep learning techniques for algebraic topology, and especially Hodge numbers, of complete intersection Calabi-Yau (CICY) manifolds [10-12].