Mathematical Physics and String Theory
print


Breadcrumb Navigation


Content

(25.01.) Branes with Brains - Reinforcement learning in the landscape of intersecting brane worlds

Fabian Ruehle (Oxford)

25.01.2018 at 16:15

We apply machine learning to an MSSM search in the string landscape of Type II intersecting branes on orientifolds of toroidal orbifolds. The advantage of these string models is that they are well understood; they have been analyzed extensively ten years ago by Ralph Blumenhagen and his group in Munich. They estimated the landscape to contain one MSSM-like model in a billion models. These large numbers make this setup an interesting case study for the applicability of machine learning to study the string landscape. I will provide a pedagogical introduction to machine learning in general, with particular emphasis on the subfield of reinforcement learning. After that, I will explain how to map the string problem to an environment that can be analysed in reinforcement learning. I will end with preliminary results of our Type II string landscape analysis.

Arnold Sommerfeld Center
Theresienstrasse 37
Room 348