Mini-workshop: Machine learning for experimental condensed matter physics

Workshop Image

Department of Physics, University of Zurich

The aim of this mini-workshop is to familiarize participants with the principles of neural-network-based machine-learning applications in condensed matter physics. Besides the theoretical background, we will also demonstrate elementary examples and the participants will learn how to set up a simple neural network calculation in the google-developed package tensor flow. No previous knowledge is required — except for python, in order to participate in the hands-on programming session. The topics are geared towards data analysis, not so much in the direction of the more fancy applications of neural networks in theoretical condensed matter physics.

Program:
Titus Neupert (Slides (PDF, 23 MB))
Lecture I: Basics of neural networks
Lecture II: Overview of neural network applications in condensed matter physics

Kenny Choo and Mark Fischer: Demonstration examples (Notebook)
Eliska Greplova: Flake searching with AI (Slides (PDF, 46 MB)Notebook)
Frank Schindler: hands-on exercise session (Notebook)

Inquiries

For inquiries, please contact frank@physik.uzh.ch.