MLQE2022

Welcome to Machine Learning for Quantum Experiments 2022

This site is an up-to-date resource for the course. All the learning materials and important updates will be published here on a daily basis.

To participate in the course, please bring a laptop. We will be using online teaching environment so installation of no specific packages beforehand is needed. For formalities and GS credits please check the official graduate school website for the course here.

This three day block course is taking place October 26-28 (10am - 5pm), on-site at TU Delft at the following locations:

Day 1: C1.190 (Building 58, van der Maasweg 9)

Day 2: C1.190 (Building 58, van der Maasweg 9)

Day 3: Room 0.96 (Building 23, Stevinweg 1)

Schedule

Materials


NOTEBOOKS DAY 1: Notebook 1, Notebook 2

NOTEBOOKS DAY 2: Notebook 3, Notebook 4

NOTEBOOKS DAY 3: Notebook 5, Notebook 6

BONUS MATERIAL: Functions and Classes in Python Notebook, Unsupervised Learning Notebook

SLIDES:

Slides Day 1: Neural Networks and Supervised Learning in Spin System

Bonus Slides Day 1: Functions and Classes in Python

Slides Day 2: Automated Tuning of Quantum Devices with Neural Networks

Slides Day 3: Quantum Parameter Estimationn and Advanced Methods


You finished MLQE2022 - what to do next?

The best advice with ML is to learn by doing, try to complete all the course notebooks and really understand the code. PyTorch website has a great tutorials and documentation that will help you to learn more. Best practice is to find a small problem you want to solve with ML and try to write a code that does so from scratch.

On the content and theory side, these are some of our favorite resources:

  • first let me advertise a Jupyterbook we wrote with colleagues from Zurich, it contains more details for most things I discussed in the lecture and lots of extra exercises

  • wonderful book by Michael Nielsen on Neural Networks and Deep Learning is a place to head to for deeper conceptual understanding

  • Deep Learning for Computer Vision is a course ran by Fei-Fei Li at Stanford and has the best no-nonsense explanation of backpropagation, convolutions and neural net design

  • For relating ML to physics, you can check out arXiv pre-print of another book we wrote with colleagues on Modern Applications of Machine Learning in Quantum Physics and Chemistry

  • Another great ML for physics lectures are these by Florian Marquardt

  • If you were interested in some of the QMAI research topics I mentioned during this course, you can check out our papers or watch some of my lectures