theoretical ML Paper Discussion Group (theory-PDG)

We meet weekly to discuss theoretical machine learning related papers and methods. The setup is whiteboard session and small group style. The most important rule: If we cannot derive the thoughts in the paper on the board ourselves, we do not understand them.

Future of Theory-PDG (09. October 2018 until X)

• Bayesian NN
• Statistical Learning Theory

Reading History

• Finished 02. October 2018 - Fisher-Rao Metric, Geometry, and Complexity of Neural Networks https://arxiv.org/abs/1711.01530
  • Notes Summary
  • some fun experiments
  • talk for other ML-KA members