Interpretable Deep Learning for New Physics Discovery

In this video, Miles Cranmer discusses a method for converting a neural network into an analytic equation using a particular set of inductive biases. The technique relies on a sparsification of latent spaces in a deep neural network, followed by symbolic regression. In their paper, they demonstrate that they can recover physical laws for various simple and complex systems. For example, they discover gravity along with planetary masses from data; they learn a technique for doing cosmology with cosmic voids and dark matter halos; and they show how to extract the Euler equation from a graph neural network trained on turbulence data.

Paper: https://arxiv.org/abs/2006.11287
Blogpost: https://astroautomata.com/paper/symbolic-neural-nets/
Code and interactive demo: https://github.com/MilesCranmer/symbolic_deep_learning
Personal website: https://astroautomata.com/
Twitter: @MilesCranmer

%%% CHAPTERS %%%
0:00 Introduction
1:51 Symbolic Regression Intro
5:38 Genetic Algorithms for Symbolic Regression
7:49 PySR for Symbolic Regression
9:09 Combining Deep Learning and Symbolic Regression
14:05 Graph Neural Networks
16:39 Recovering Physics from a GNN
18:39 Results on Unknown Systems
23:12 Takeaways

Видео Interpretable Deep Learning for New Physics Discovery канала Steve Brunton