Close

Filter courses

Show courses Show filters
Close

Machine Learning for Semiconductor Quantum Devices

01 /Supervised Learning for quantum Dots 02 /Charge Stability Diagrams 03 /Classify CS Diagrams with Neural Networks 04 /Supervised Learning with Feedback 05 /Charge Tuning as a NN with Feedback 06 /Real-Life Experimental Applications 07 /Unsupervised Learning for Quantum Dots 08 /Analyse CS Diagrams with Clustering 09 /Analyse CS Diagrams with Unsupervised NNs 10 /Autoencoders in QD Experiments 11 /ML Fine Tuning of Quantum Dots 12 /Overhauser Field estimation 13 /Recap 14 /State of the Art Tuning

Analyse CS Diagrams with Clustering

In this video we explain how to employ Principal Component Analysis to recover data patterns in Charge Stability Diagrams. We discuss the interpretation of the principal components.

Main takeaways

  • Quite simple linear clustering techniques can be enormously useful in recovering features in a dataset
  • In the case of unsupervised learning, one must be careful about interpreting the clusters of features correctly. Just an existence of a cluster itself does not imply the cluster corresponds to the physics we expect.

Further thinking

In principal component analysis, what are the principal components?

a. vectors that minimize the square distance of the data
b. vectors that maximize the square distance of the data
c. dataset samples with the lowest variance
d. dataset features with the lowest variance

Solution

Further reading

Discovering Phase Transitions with Unsupervised Learning: https://arxiv.org/pdf/1606.00318.pdf

Paper: Quantum device fine-tuning using unsupervised embedding learning: https://iopscience.iop.org/article/10.1088/1367-2630/abb64c

Efficiently measuring a quantum device using machine learning: https://www.nature.com/articles/s41534-019-0193-4