Machine Learning and the Real Space Renormalization Group

Explore the intersection of machine learning and the Renormalization Group in this 30-minute Physics Next talk from 2018. Delve into the Real-space Renormalization Group from an Information Theory perspective, examining concepts such as Mutual Information and their applications. Analyze test cases including the 2D Ising model and the dimer model, while investigating RG flow and critical exponents. Gain insights into the optimality of the Real Space Mutual Information (RSMI) approach as presented by Maciej Kock-Janusz from the Swiss Federal Institute of Technology in Zurich.

Intro
Outline
Renormalization Group
Real-space RG from Information Theory perspective
Mutual Information
Test case 1: the 2D Ising model
RG flow and critical exponents
Test case 2: the dimer model
Optimality of the RSMI approach