Machine Learning for Quantum Technology
Machine Learning for Quantum Technology, Erlangen 2019
I had a chance to attend this workshop last month, it was a short but quite intense overview of applications of standard machine learning techniques to various problems in the quantum world. It was an interesting perspective to see, both from a theoretical and an experimental point of view, applications of some well known machine learning models in a context that was not the MNIST or CIFAR10 benchmarks and to see some real-world usage of it. A notable example is the work done by Natalia Ares and collaborators, they fabricate quantum dots and, in order to test and tune them, they have to take a lot of measurements often with random parameters just to find the “sweet spot” of the device. They successfully created a model to infer an “information gain” map from the data and thus optimize the measure they have to take in order to correctly characterize their devices. The code is open sourced and you can find it here:
https://github.com/returnddd/CVAE_for_QE
This workshop provided also a good introduction of “machine learning for physicists” that was a good example of bridging the language gap that often exists between the two disciplines. I highly recommend whoever might be interest to check out the site of the workshop where most of the presentation have been uploaded:
Hope that this brief summary was as interesting as it was the workshop itself for me.
Lorenzo Buffoni