A Robust and Efficient Deep Learning Method for Dynamical Mass Measurements of Galaxy Clusters

Journal Article · 06 December 2019 · The Astrophysical Journal (Online)

DOI: https://doi.org/10.3847/1538-4357/ab4f82 · OSTI ID:1802212

Ho, Matthew  ^[1]; Rau, Markus Michael  ^[2]; Ntampaka, Michelle  ^[3]; Farahi, Arya  ^[2]; Trac, Hy  ^[2]; Póczos, Barnabás ^[4]

---

1. Carnegie Mellon Univ., Pittsburgh, PA (United States). McWilliams Center for Cosmology, Dept. of Physics; Carnegie Mellon University
2. Carnegie Mellon Univ., Pittsburgh, PA (United States). McWilliams Center for Cosmology, Dept. of Physics
3. Harvard Univ., Cambridge, MA (United States). Harvard Data Science Initiative; Harvard-Smithsonian Center for Astrophysics, Cambridge, MA (United States)
4. Carnegie Mellon Univ., Pittsburgh, PA (United States). School of Computer Science

We demonstrate the ability of convolutional neural networks (CNNs) to mitigate systematics in the virial scaling relation and produce dynamical mass estimates of galaxy clusters with remarkably low bias and scatter. We present two models, CNN_1D and CNN_2D, which leverage this deep learning tool to infer cluster masses from distributions of member galaxy dynamics. Our first model, CNN_1D, infers cluster mass directly from the distribution of member galaxy line-of-sight velocities. Our second model, CNN_2D, extends the input space ofCNN_1D to learn on the joint distribution of galaxy line-of-sight velocities and projected radial distances. We train each model as a regression over cluster mass using a labeled catalog of realistic mock cluster observations generated from the MultiDark simulation and UniverseMachine catalog. We then evaluate the performance of each model on an independent set of mock observations selected from the same simulated catalog. The CNN models produce cluster mass predictions with lognormal residuals of scatter as low as 0.132 dex, greater than a factor of 2 improvement over the classical M–σ power-law estimator. Furthermore, the CNN model reduces prediction scatter relative to similar machinelearning approaches by up to 17% while executing in drastically shorter training and evaluation times (by a factor of 30) and producing considerably more robust mass predictions (improving prediction stability under variations in galaxy sampling rate by 30%).

• Research Organization: Carnegie Mellon Univ., Pittsburgh, PA (United States)
• Sponsoring Organization: USDOE Office of Science (SC)
• Grant/Contract Number: SC0011114
• OSTI ID: 1802212
• Journal Information: The Astrophysical Journal (Online), Journal Name: The Astrophysical Journal (Online) Journal Issue: 1 Vol. 887; ISSN 1538-4357
• Publisher: IOP PublishingCopyright Statement
• Country of Publication: United States
• Language: English