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.
- Bayesian NN
- Statistical Learning Theory
- 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