December 13, 2021 12:02

We are glad to inform you that the Approximate Bayesian Inference (ABI) team won the NeurIPS challenge on “Approximate Inference in Deep Learning”. The team comprised of the following individuals,

  • Thomas Möllenhoff (ABI, RIKEN-AIP)
  • Yuesong Shen (TU Munich)
  • Gian Maria Marconi (ABI, RIKEN-AIP)
  • Peter Nickl (ABI, RIKEN-AIP)
  • Mohammad Emtiyaz Khan (ABI, RIKEN-AIP)

Thomas Möllenhoff, Postdoctoral Researcher of the ABI delivered the results at a NeurIPS event on Dec. 9, 6 pm GMT, and also he will give a talk at the Bayesian deep learning workshop on Dec 14, 4:40 pm GMT.

The methods used in the winning entry were based on the following two papers.

1. The Bayesian Learning Rule, M.E. Khan, H. Rue [ arXiv ]
2. Handling the Positive-Definite Constraint in the Bayesian Learning Rule, (ICML 2020) W. Lin, M. Schmidt, M.E. Khan [ arXiv ]

Related Laboratories

last updated on December 13, 2021 10:30Laboratory