As supernova surveys become larger and more efficient, photometric supernova cosmology will become the prevalent route to make the best use of the large data sets. However, photometric typing is unreliable, and a non-negligible fraction of the 'wrong' kind of supernovae will bias the inference of cosmological parameters. To deal with this problem we have developed the Bayesian Estimation Applied to Multiple Species (BEAMS) formalism which employs a mixture model in a Bayesian context to let the data decide about the nature of the supernovae. I will discuss the statistical basis of the algorithm and its application to actual supernova data, and show how it enables the use of larger supernova samples for science.
Keywords: Astronomy; Supernovae; Cosmology; Bayesian statistics
Biography: Dr Kunz is a faculty member at the Institute of Theoretical Physics of the University of Geneva, he has previously been working at the Universities of Sussex and Oxford. His research focuses on theoretical and statistical challenges in cosmology, especially in the context of the dark energy problem.