In studying the vacua of string theory, researchers often want to find solutions with specific properties, but do not know how to select the string geometry that gives rise to such vacua. For this reason, the speaker applies reinforcement learning, a semi-supervised approach to machine learning, in which the algorithm explores the landscape of string solutions autonomously while being guided towards models with given properties. In this talk, he will give an introduction to reinforcement learning and illustrate the mechanism in several examples from type II string theory, heterotic string theory, and F-theory.
