Exploring photometric redshifts as an optimization problem: an ensemble MCMC and simulated annealing-driven template-fitting approach
Abstract
Using a 4D grid of ∼2 million model parameters (Δz = 0.005) adapted from Cosmological Origins Survey photometric redshift (photo-z) searches, we investigate the general properties of template-based photo-z likelihood surfaces. We find these surfaces are filled with numerous local minima and large degeneracies that generally confound simplistic gradient-descent optimization schemes. We combine ensemble Markov Chain Monte Carlo sampling with simulated annealing to robustly and efficiently explore these surfaces in approximately constant time. Using a mock catalogue of 384 662 objects, we show our approach samples ∼40 times more efficiently compared to a 'brute-force' counterpart while maintaining similar levels of accuracy. Our results represent first steps towards designing template-fitting photo-z approaches limited mainly by memory constraints rather than computation time.
Additional Information
© 2016 The Authors Published by Oxford University Press on behalf of the Royal Astronomical Society. Accepted 2016 June 21. Received 2016 June 17. In original form 2015 August 10. First published online June 24, 2016. The authors would like to thank the anonymous referee for useful comments that substantially improved the quality of this work. JSS would like to thank Douglas Finkbeiner and Zachary Slepian for comments that improved the quality of this work and Charles Alcock for supervising the senior thesis course where much of this work was completed. JSS is grateful for financial support from the Herchel Smith-Harvard Summer Undergraduate Research Fellowship, the Harvard University Department of Astronomy, the Harvard College Observatory, and CREST funding from the Japan Science and Technology Agency (JST). This work has benefited extensively from access to computing resources at IPAC and Harvard University.Attached Files
Published - MNRASSpeagle,JSetal.pdf
Submitted - 1508.02484v1.pdf
Files
Name | Size | Download all |
---|---|---|
md5:35806ff94b0f19accad01c9d80c0eb92
|
4.5 MB | Preview Download |
md5:065833e041e680e8c9697b6c75b663e1
|
2.1 MB | Preview Download |
Additional details
- Eprint ID
- 71179
- Resolver ID
- CaltechAUTHORS:20161017-130307346
- Harvard University
- Harvard College Observatory
- Japan Science and Technology Agency (JST)
- Created
-
2016-10-17Created from EPrint's datestamp field
- Updated
-
2021-11-11Created from EPrint's last_modified field
- Caltech groups
- Infrared Processing and Analysis Center (IPAC)