Revealing the Structure of Data in Physics by Machine Learning
Abstract: In this talk, I will introduce machine-learning methods for discovering and exploiting the structure of data in physics. In the first part, we study the automatic detection of phase transitions, which is the classification of data in a phase diagram. The fact that these data are associated with continuous tuning parameters allows for powerful new algorithms. In the second part, we study machine-learning aided quantum error correction, where the power of neural networks is utilized to fight the noise in a transmission channel.
1. Nat. Phys. 13, 435 (2017).
2. Phys. Rev. Lett. 120, 176401 (2018).