Physically constrained causal noise models for high-contrast imaging of exoplanets

Timothy D. Gebhard, Markus J. Bonse, Sascha P. Quanz, Bernhard Schölkopf

Accepted at the Machine Learning and the Physical Sciences workshop at NeurIPS 2020, October 2020

Cite this paper

@article{Gebhard_2020,
  title         = {{Physically constrained causal noise models for high-contrast imaging of exoplanets}},
  author        = {Timothy D. Gebhard and Markus J. Bonse and Sascha P. Quanz and Bernhard Schölkopf},
  year          = 2020,
  month         = 10,
  eprint        = {2010.05591},
  eprinttype    = {arXiv},
  addendum      = {Accepted at the Machine Learning and the Physical Sciences workshop at NeurIPS 2020},
}

NASA/ADS

Abstract:

The detection of exoplanets in high-contrast imaging (HCI) data hinges on post-processing methods to remove spurious light from the host star. So far, existing methods for this task hardly utilize any of the available domain knowledge about the problem explicitly. We propose a new approach to HCI post-processing based on a modified half-sibling regression scheme, and show how we use this framework to combine machine learning with existing scientific domain knowledge. On three real data sets, we demonstrate that the resulting system performs up to a factor of 4 times better than one of the currently leading algorithms. This has the potential to allow significant discoveries of exoplanets both in new and archival data.