The Application Of Deep Learning In Holographic Gravity Model Construction

As the holographic property of gravity is discovered,people have a new understanding of gravity and space-time.Holographic theory is expected to be used to coordinate the contradiction between quantum mechanics and general relativity,thus opening a door of quantum gravity.As an example of holographic principle,AdS/CFT duality also has important applications in condensed matter physics,particle physics and other fields.Meanwhile,as an important branch of computer science,deep learning has developed rapidly in recent years.With the great improvement of today's computer operation ability,coupled with the data explosion in the information era,the rapid rise of deep learning is slowly exerting profound impact on human society.Using deep learning as a powerful tool to study holographic theory is a very significant topic.This thesis uses the Reissner-Nordstr?m black hole as the background to construct the AdS model by constructing a deep neural network,and uses the deep learning approach to learn the information on the AdS boundary,after learning the boundary information for a period of time,the deep neural network gives the whole metric of Reissner-Nordstr?m black hole.It is good enough to show us the feasibility of deep learning in studying holographic theory.In addition,we delved into the effects of various parameters in deep neural networks on learning outcomes.