Astrophysics > Cosmology and Nongalactic Astrophysics
[Submitted on 19 Oct 2018]
Title:An application of machine learning techniques to galaxy cluster mass estimation using the MACSIS simulations
View PDFAbstract:Machine learning (ML) techniques, in particular supervised regression algorithms, are a promising new way to use multiple observables to predict a cluster's mass or other key features. To investigate this approach we use the \textsc{MACSIS} sample of simulated hydrodynamical galaxy clusters to train a variety of ML models, mimicking different datasets. We find that compared to predicting the cluster mass from the $\sigma -M$ relation, the scatter in the predicted-to-true mass ratio is reduced by a factor of 4, from $0.130\pm0.004$ dex (${\simeq} 35$ per cent) to $0.031 \pm 0.001$ dex (${\simeq} 7$ per cent) when using the same, interloper contaminated, spectroscopic galaxy sample. Interestingly, omitting line-of-sight galaxy velocities from the training set has no effect on the scatter when the galaxies are taken from within $r_{200c}$. We also train ML models to reproduce estimated masses derived from mock X-ray and weak lensing analyses. While the weak lensing masses can be recovered with a similar scatter to that when training on the true mass, the hydrostatic mass suffers from significantly higher scatter of ${\simeq} 0.13$ dex (${\simeq} 35$ per cent). Training models using dark matter only simulations does not significantly increase the scatter in predicted cluster mass compared to training on simulated clusters with hydrodynamics. In summary, we find ML techniques to offer a powerful method to predict masses for large samples of clusters, a vital requirement for cosmological analysis with future surveys.
Current browse context:
astro-ph.GA
Change to browse by:
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
IArxiv Recommender
(What is IArxiv?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.