% IMPORTANT: The following is UTF-8 encoded.  This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.

@INPROCEEDINGS{Bauer:874367,
      author       = {Bauer, Carsten and Trebst, Simon},
      title        = {{M}achine {L}earning {T}ransport {P}roperties in {Q}uantum
                      {M}any-{F}ermion {S}imulations},
      volume       = {50},
      address      = {Jülich},
      publisher    = {Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag},
      reportid     = {FZJ-2020-01390},
      series       = {Publication Series of the John von Neumann Institute for
                      Computing (NIC) NIC Series},
      pages        = {85 - 92},
      year         = {2020},
      comment      = {NIC Symposium 2020},
      booktitle     = {NIC Symposium 2020},
      abstract     = {In computational condensed matter physics, the influx of
                      algorithms from machine learning and their combination with
                      traditional numerical many-body approaches is one of the
                      most enticing recent developments. At this confluence novel
                      techniques have been developed that allow to characterise
                      many-body wave functions and discriminate quantum phase of
                      matter by adapting concepts from computer science and
                      statistics, which have proved tremendously practical in
                      completely different contexts. However, in order to actually
                      turn into a productive and widely accepted tool for
                      obtaining a deeper understanding of microscopic physics
                      these novel approaches must allow for meaningful,
                      comprehensible inference and go beyond the applicability of
                      their traditional counterparts. In this contribution, we
                      report on significant progress made in this direction by
                      discussing a novel algorithmic scheme using machine learning
                      techniques to numerically infer the transport properties of
                      quantum many-fermion systems. This approach is based on a
                      quantum loop topography (QLT), and capable of distinguishing
                      conventional metallic and superconducting transport in
                      quantum Monte Carlo simulations by learning current-current
                      correlations from equal-time Green’s functions. We
                      showcase this approach by studying the emergence of s- and
                      d-wave superconducting fluctuations in the negative-U
                      Hubbard model and a spin-fermion model for a metallic
                      quantum critical point. The presented results, combined with
                      the numerical efficiency of the QLT approach, point a way to
                      identify hitherto elusive transport phenomena such as
                      non-Fermi liquids using machine learning algorithms.},
      month         = {Feb},
      date          = {2020-02-27},
      organization  = {NIC Symposium 2020, Jülich (Germany),
                       27 Feb 2020 - 28 Feb 2020},
      cin          = {NIC},
      cid          = {I:(DE-Juel1)NIC-20090406},
      pnm          = {899 - ohne Topic (POF3-899)},
      pid          = {G:(DE-HGF)POF3-899},
      typ          = {PUB:(DE-HGF)8 / PUB:(DE-HGF)7},
      url          = {https://juser.fz-juelich.de/record/874367},
}