CANCELLED! QOPT seminar by Raffaele Santagati University of Bristol

Experimentally learning the properties of quantum systems with machine learning 

Abstract:The characterisation of quantum systems is of paramount importance for deepening our understanding of their physics and for the development of quantum technologies [1]. However, quantum systems are known for being challenging to characterise, and it is also often difficult to synthesise good predictive models with clear physical interpretations. 

Quantum Hamiltonian learning was proposed as a solution to the problem of estimating Hamiltonian parameters [2], making more practical the characterisation of quantum systems. 

In the first part of this talk, I will present how quantum Hamiltonian learning protocols can find a practical application on quantum sensing experiments with NV centres in diamond. By transforming the magnetic field measurement into a Hamiltonian parameter estimation problem we can achieve at room temperature performance typical of cryogenic set-ups [3]. 

In the second part, I will show how Hamiltonian learning can be generalised to the case of an unsupervised machine learning protocol able to synthesise Hamiltonian models of solid-state quantum systems from experimental observations [4].

[1] Da Silva, et al. P.R.L. 107, 210404, 2011 [2] Wiebe, et al. P.R.L. 112, 190501, 2014  [3] Santagati, et al. P.R.X. 9, 021019, 2019 [4] Gentile, et al. arXiv:2002.06169, 2019