Deep learning the nonclassicality within quasi-distribution representations from marginals

  • Event Date: 2023-09-11
  • Quantum information and communication
  • Speaker: Prof. Hong-Bin Chen(Dept. of Engineering Science, NCKU)  /  Host: Prof. Yueh-Nan Chen(NCKU)
    Place: R36173, 1F, Dept. of Physics, Building of Science College, NCKU

Time:12:10, Monday, September 11, 2023
Speaker:Prof. Hong-Bin Chen
              Dept. of Engineering Science, NCKU
Title: Deep learning the nonclassicality within quasi-distribution representations from marginals
Place : R36173, 1F, Dept. of Physics, Building of Science College, NCKU

Abstract:
To unequivocally distinguish the genuine quantumness from classicality, a widely adopted approach appeals to the negativity within a join quasi-distribution representation as a compelling evidence for the nonclassical essence. However to construct a joint quasi-distribution with negativity from experimental data typically proves to be highly cumbersome. Here we design a deep generative model integrated with color mapping to construct the bivariate joint quasi-distribution functions by processing three marginals. Our model successfully predicts both the Wigner functions and the recently developed canonical Hamiltonian ensemble representation (CHER) characterizing the dynamical process nonclassicality. By processing three marginals of probability distributions, our model accurately predicts the Wigner functions. Furthermore, we also design optimal synthetic datasets to train the model for overcoming the GT-deficiency of the CHER problem. While trained with synthetic data, the physics-informed optimization enables our model to capture the detrimental effect of the thermal fluctuations on nonclassicality. Our approach also provides a significant reduction of the experimental efforts of constructing the Wigner functions of quantum states.