The 1st Workshop on Deep Continuous-Discrete Machine Learning will take place on Monday, September 16, 2019, in W├╝rzburg Germany, associated to ECML PKDD 2019.

The workshop intends to attract papers on how recent deep learning methods can be connected to discrete structures and symbolic knowledge, including areas such as

  • Pedagogical, decompositional, or hybrid pedagogical/decompositional knowledge/rule extraction from deep neural networks
  • Prior knowledge used as constraints in deep neural networks (e.g., compiled into the architecture/topology, taking advantage of known symmetries, ...)
  • Deep symbolic networks, e.g., networks of rules or trees, deep rule sets
  • Deep hybrid continuous-discrete networks
  • HPC or GPU implementations
  • Applications to text, images, sequences, time series, ...

Important Dates and procedures: