Dark Matter Subhalos, Strong Lensing and Machine Learning

Abstract

We investigate the possibility of applying machine learning techniques to images of strongly lensed galaxies to detect a low mass cut-off in the spectrum of dark matter sub-halos within the lens system. We generate lensed images of systems containing substructure in seven different categories corresponding to lower mass cut-offs ranging from $10^9{M}_{\odot }$ down to $10^6{M}_{\odot }$. We use convolutional neural networks to perform a multi-classification sorting of these images and see that the algorithm is able to correctly identify the lower mass cut-off within an order of magnitude to better than 93% accuracy.

20 pages

References in corpus (57)

• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
• ADADELTA: An Adaptive Learning Rate Method
• The clustering of galaxies in the completed SDSS-III Baryon Oscillation Spectroscopic Survey: cosmological analysis of the DR12 galaxy sample
• The Complete Light-curve Sample of Spectroscopically Confirmed Type Ia Supernovae from Pan-STARRS1 and Cosmological Constraints from The Combined Pantheon Sample
• Machine learning and the physical sciences
• The Aquarius Project: the subhalos of galactic halos
• Dark Matter Search Results from a One Tonne$\times$Year Exposure of XENON1T
• Cold and Fuzzy Dark Matter
• Ultralight scalars as cosmological dark matter
• Results from a search for dark matter in the complete LUX exposure
• Small-Scale Challenges to the $\Lambda$CDM Paradigm
• Sterile Neutrinos as Dark Matter
• Dark Matter Self-interactions and Small Scale Structure
• Cosmological Simulations with Self-Interacting Dark Matter I: Constant Density Cores and Substructure
• The Milky Way's bright satellites as an apparent failure of LCDM
• Earth-mass dark-matter haloes as the first structures in the early Universe
• Compound gravitational lensing as a probe of dark matter substructure within galaxy halos
• Lenstronomy: multi-purpose gravitational lens modelling software package
• Bayesian Strong Gravitational-Lens Modeling on Adaptive Grids: Objective Detection of Mass Substructure in Galaxies
• Fast Automated Analysis of Strong Gravitational Lenses with Convolutional Neural Networks
• The accretion history of dark matter halos III: A physical model for the concentration-mass relation
• Dark matter heats up in dwarf galaxies
• Possible dark matter annihilation signal in the AMS-02 antiproton data
• SHARP -- VII. New constraints on the dark matter free-streaming properties and substructure abundance from gravitationally lensed quasars
• The chosen few: the low mass halos that host faint galaxies
• Search for dijet resonances in proton-proton collisions at sqrt(s) = 13 TeV and constraints on dark matter and other models
• Uncertainties in Parameters Estimated with Neural Networks: Application to Strong Gravitational Lensing
• Linear perturbation theory for tidal streams and the small-scale CDM power spectrum
• The Substructure Hierarchy in Dark Matter Haloes
• Lensing substructure quantification in RXJ1131-1231: A 2 keV lower bound on dark matter thermal relic mass
• The impact of baryonic physics on the subhalo mass function and implications for gravitational lensing
• Measuring the power spectrum of dark matter substructure using strong gravitational lensing
• The core-cusp problem: a matter of perspective
• Substructure of fuzzy dark matter haloes
• Constraints on mediator-based dark matter and scalar dark energy models using $\sqrt{s}=13$ TeV $pp$ collision data collected by the ATLAS detector
• On the dark matter profile in Sculptor: Breaking the $\beta$ degeneracy with Virial shape parameters
• Deep Convolutional Neural Networks as strong gravitational lens detectors
• Evidence of a population of dark subhalos from Gaia and Pan-STARRS observations of the GD-1 stream
• Abundance matching with the mean star formation rate: there is no missing satellites problem in the Milky Way above ${\mathbf{M}}_{200}\sim 10^9{\mathbf{M}}_{\odot }$
• Identifying Strong Lenses with Unsupervised Machine Learning using Convolutional Autoencoder
• Mining for Dark Matter Substructure: Inferring subhalo population properties from strong lenses with machine learning
• MeV Dark Matter and Small Scale Structure
• A Search for Dark Matter Annihilation in Galaxy Groups
• Using Convolutional Neural Networks to identify Gravitational Lenses in Astronomical images
• To beta or not to beta: can higher-order Jeans analysis break the mass-anisotropy degeneracy in simulated dwarfs?
• On the Power Spectrum of Dark Matter Substructure in Strong Gravitational Lenses
• Gravitational Lensing and the Power Spectrum of Dark Matter Substructure: Insights from the ETHOS N-body Simulations
• A Search for Dark Matter Annihilation in the Milky Way Halo
• A Dark Census: Statistically Detecting the Satellite Populations of Distant Galaxies
• Direct Detection of Dark Matter Substructure in Strong Lens Images with Convolutional Neural Networks
• Probing the small-scale structure in strongly lensed systems via transdimensional inference
• Differentiable Strong Lensing: Uniting Gravity and Neural Nets through Differentiable Probabilistic Programming
• Predicting dark matter halo formation in N-body simulations with deep regression networks
• Analyzing interferometric observations of strong gravitational lenses with recurrent and convolutional neural networks
• Quantifying the power spectrum of small-scale structure in semi-analytic galaxies
• Gravitational Imaging of CDM Substructure
• Hunt for dark subhalos in the galactic stellar field using computer vision

Cited by in corpus (4)

• Decoding Dark Matter Substructure without Supervision
• An Accurate Comprehensive Approach to Substructure: II. Stripped Subhaloes
• Towards a method to anticipate dark matter signals with deep learning at the LHC
• Embedded-model flows: Combining the inductive biases of model-free deep learning and explicit probabilistic modeling