Speaker
Igor Moskalenko
(Stanford University)
Description
The fully Bayesian approach to the problem of deriving constraints for cosmic ray (CR) model parameters has several advantages. These are: (i) an efficient global scan of the whole parameter space allowing us to explore and take into account parameter correlations and degeneracies, (ii) a best-fit point and statistically well-defined errors on the parameters, (iii) the ability to include and marginalize over "nuisance" parameters (such as modulation potential and error rescaling parameters) making the analysis more robust. For this study, we use the latest version of the CR propagation code GALPROP together with the BAMBI code, the most efficient Bayesian analysis code available to date that combines MultiNest with Neural networks. The results of the analysis will be reported during the conference.
| Registration number following "ICRC2015-I/" | 382 |
|---|---|
| Collaboration | -- not specified -- |
Authors
Aaron Vincent
(IPPP, Durham University)
Gudlaugur Johannesson
(Science Institute, University of Iceland)
Roberto Ruiz De Austri
(IFIC)
Co-authors
Andrew Strong
(MPE fuer extraterrestrische Physik)
Elena Orlando
(Stanford University)
Igor Moskalenko
(Stanford University)
Roberto Trotta
(Imperial College London)
Troy Porter
(Stanford University)
