Approximate Bayesian Computation for Astronomical Model Analysis: A Case Study in Galaxy Demographics and Morphological Transformation at High Redshift

Abstract

"Approximate Bayesian Computation" (ABC) represents a powerful methodology for the analysis of complex stochastic systems for which the likelihood of the observed data under an arbitrary set of input parameters may be entirely intractable-the latter condition rendering useless the standard machinery of tractable likelihood-based, Bayesian statistical inference (e.g. conventional Markov Chain Monte Carlo simulation; MCMC). In this article we demonstrate the potential of ABC for astronomical model analysis by application to a case study in the morphological transformation of high redshift galaxies. To this end we develop, first, a stochastic model for the competing processes of merging and secular evolution in the early Universe; and second, through an ABC-based comparison against the observed demographics of massive (M_gal > 10^11 M_sun) galaxies (at 1.5 < z < 3) in the CANDELS/EGS dataset we derive posterior probability densities for the key parameters of this model. The "Sequential Monte Carlo" (SMC) implementation of ABC exhibited herein, featuring both a self-generating target sequence and self-refining MCMC kernel, is amongst the most efficient of contemporary approaches to this important statistical algorithm. We highlight as well through our chosen case study the value of careful summary statistic selection, and demonstrate two modern strategies for assessment and optimisation in this regard. Ultimately, our ABC analysis of the high redshift morphological mix returns tight constraints on the evolving merger rate in the early Universe and favours major merging (with disc survival or rapid reformation) over secular evolution as the mechanism most responsible for building up the first generation of bulges in early-type disks.

25 pages, 11 figures, submitted to MNRAS, revised once

References in corpus (47)

• The hierarchical formation of the brightest cluster galaxies
• Formation of Massive Galaxies at High Redshift: Cold Streams, Clumpy Disks and Compact Spheroids
• Galaxy And Mass Assembly: Stellar Mass Estimates
• The All-wavelength Extended Groth Strip International Survey (AEGIS) Data Sets
• A primer on hierarchical galaxy formation: the semi-analytical approach
• Rapid formation of exponential disks and bulges at high redshift from the dynamical evolution of clump cluster and chain galaxies
• Can Minor Merging Account for the Size Growth of Quiescent Galaxies? New Results from the CANDELS Survey
• Galaxy Merger Morphologies and Time-Scales from Simulations of Equal-Mass Gas-Rich Disc Mergers
• CANDELS: Constraining the AGN-Merger Connection with Host Morphologies at z~2
• Bulge Formation by the Coalescence of Giant Clumps in Primordial Disk Galaxies
• Galaxy And Mass Assembly (GAMA): Structural Investigation of Galaxies via Model Analysis (SIGMA)
• What turns galaxies off? The different morphologies of star-forming and quiescent galaxies since z~2 from CANDELS
• Generating Dark Matter Halo Merger Trees
• A massive proto-cluster of galaxies at a redshift of z {\approx} 5.3
• Sizes and surface brightness profiles of quiescent galaxies at z ~ 2
• HST WFC3/IR Observations of Active Galactic Nucleus Host Galaxies at z~2: Supermassive Black Holes Grow in Disk Galaxies
• The Structures of Distant Galaxies - III: The Merger History of over 20,000 Massive Galaxies at z < 1.2
• The structures and total (minor + major) merger histories of massive galaxies up to z = 3 in the HST GOODS NICMOS Survey: A possible solution to the size evolution problem
• Early type galaxies have been the predominant morphological class for massive galaxies since only z~1
• CANDELS Observations of the Structural Properties and Evolution of Galaxies in a Cluster at z=1.62
• Stellar Feedback & Bulge Formation in Clumpy Disks
• First Results from Z-FOURGE: Discovery of a Candidate Cluster at z = 2.2 in COSMOS
• The Bright End of the UV Luminosity Function at z~8: New Constraints from CANDELS Data in GOODS-South
• MOIRCS Deep Survey IV: Evolution of Galaxy Stellar Mass Function Back to z ~ 3
• An Improved Model of SiO Maser Emission in Miras
• LCDM predictions for galaxy protoclusters I: the relation between galaxies, protoclusters and quasars at z~6
• The Pair Fraction of Massive Galaxies at 0 < z < 3
• Estimation of cosmological parameters using adaptive importance sampling
• Monte Carlo Markov Chain Parameter Estimation in Semi-Analytic Models of Galaxy Formation
• Morphological evolution of galaxies from ultradeep HST WFC3 imaging: the Hubble sequence at z~2
• A Surprisingly High Pair Fraction for Extremely Massive Galaxies at z ~ 3 in the GOODS NICMOS Survey
• A Stochastic Process Approach of the Drake Equation Parameters
• Testing the predictions of the cold dark matter model for the sizes, colours, morphologies and luminosities of galaxies with the SDSS
• The Universal Initial Mass Function In The XUV Disk of M83
• Smooth and Starburst Tidal Tails in the GEMS and GOODS Fields
• The structural and photometric properties of early-type galaxies in hierarchical models
• Monolithic or hierarchical star formation? A new statistical analysis
• Environmental effects in the interaction and merging of galaxies in zCOSMOS
• Quantifying the Universality of the Stellar Initial Mass Function in Old Star Clusters
• Feedback from Clustered Sources During Reionization
• Galaxy evolution by color-log(n) type since redshift unity in the Hubble Ultra Deep Field
• The zCOSMOS redshift survey : Influence of luminosity, mass and environment on the galaxy merger rate
• Unveiling hidden properties of young star clusters: differential reddening, star-formation spread and binary fraction
• Relevant statistics for Bayesian model choice
• Approximate Bayesian computation and Bayes linear analysis: Towards high-dimensional ABC
• Probabilistic distributions of $M/L$ values for ultra-faint dwarf spheroidal galaxies: stochastic samplings of the IMF
• Caught in the Act: The Assembly of Massive Cluster Galaxies at z=1.62

Cited by in corpus (60)

• Machine learning and the physical sciences
• Fast likelihood-free cosmology with neural density estimators and active learning
• COLDz: Shape of the CO Luminosity Function at High Redshift and the Cold Gas History of the Universe
• Massive optimal data compression and density estimation for scalable, likelihood-free inference in cosmology
• Resolving the Discrepancy of Galaxy Merger Fraction Measurements at z ~ 0 - 3
• Likelihood-free inference with neural compression of DES SV weak lensing map statistics
• Nuisance hardened data compression for fast likelihood-free inference
• Variability Timescale and Spectral Index of Sgr A* in the Near Infrared: Approximate Bayesian Computation Analysis of the Variability of the Closest Supermassive Black Hole
• Automatic physical inference with information maximising neural networks
• Likelihood-Free Cosmological Inference with Type Ia Supernovae: Approximate Bayesian Computation for a Complete Treatment of Uncertainty
• Generalized massive optimal data compression
• Approximate Bayesian Computation for Forward Modeling in Cosmology
• cosmoabc: Likelihood-free inference via Population Monte Carlo Approximate Bayesian Computation
• Improving the Accuracy of Planet Occurrence Rates from Kepler using Approximate Bayesian Computation
• A new model to predict weak-lensing peak counts II. Parameter constraint strategies
• A new model to predict weak-lensing peak counts I. Comparison with $N$-body Simulations
• Simulation-Based Inference of Reionization Parameters From 3D Tomographic 21 cm Lightcone Images
• MASSIV: Mass Assembly Survey with SINFONI in VVDS. V. The major merger rate of star-forming galaxies at 0.9 < z < 1.8 from IFS-based close pairs
• Evidence of a non universal stellar Initial Mass Function. Insights from HST optical imaging of 6 Ultra Faint Dwarf Milky Way Satellites
• astroABC: An Approximate Bayesian Computation Sequential Monte Carlo sampler for cosmological parameter estimation
• What Are the Progenitors of Compact, Massive, Quiescent Galaxies at z=2.3? The Population of Massive Galaxies at z>3 from NMBS and CANDELS
• A New Method to Measure the Post-Reionization Ionizing Background from the Joint Distribution of Lyman-$\alpha$ and Lyman-$\beta$ Forest Transmission
• Approximate Bayesian Computation in Large Scale Structure: constraining the galaxy-halo connection
• Cosmological constraints with weak lensing peak counts and second-order statistics in a large-field survey
• Sequential Monte Carlo with Adaptive Weights for Approximate Bayesian Computation
• Star Formation Quenching Timescale of Central Galaxies in a Hierarchical Universe
• Implicit Likelihood Inference of Reionization Parameters from the 21 cm Power Spectrum
• Accelerated Bayesian SED Modeling using Amortized Neural Posterior Estimation
• Statistical Inference for Generative Models with Maximum Mean Discrepancy
• The sum of the masses of the Milky Way and M31: a likelihood-free inference approach
• ABC-CDE: Towards Approximate Bayesian Computation with Complex High-Dimensional Data and Limited Simulations
• A new model to predict weak-lensing peak counts III. Filtering technique comparisons
• Likelihood-free Cosmological Constraints with Artificial Neural Networks: An Application on Hubble Parameters and SNe Ia
• The PAU Survey: Measurement of Narrow-band galaxy properties with Approximate Bayesian Computation
• Quasars as probes of cosmological reionization
• Prospects for Recovering Galaxy Intrinsic Shapes from Projected Quantities
• 21st Century Statistical and Computational Challenges in Astrophysics
• Inferring the photometric and size evolution of galaxies from image simulations
• Resequencing the Hubble sequence and the quadratic (black hole mass)-(spheroid stellar mass) relation for elliptical galaxies
• CoLFI: Cosmological Likelihood-free Inference with Neural Density Estimators
• Dynamic Likelihood-free Inference via Ratio Estimation (DIRE)
• Deriving the star formation histories of galaxies from spectra with simulation-based inference
• Statistical challenges in the search for dark matter
• DR-ABC: Approximate Bayesian Computation with Kernel-Based Distribution Regression
• Machine Learning Accelerated Likelihood-Free Event Reconstruction in Dark Matter Direct Detection
• Likelihood-free Forward Modeling for Cluster Weak Lensing and Cosmology
• A Preferential Attachment Model for the Stellar Initial Mass Function
• High-Dimensional Density Ratio Estimation with Extensions to Approximate Likelihood Computation
• Test of Artificial Neural Networks in Likelihood-free Cosmological Constraints: A Comparison of IMNN and DAE
• Probably approximate Bayesian computation: nonasymptotic convergence of ABC under misspecification
• Bayesian large-scale structure inference and cosmic web analysis
• Quantifying the Projected Suppression of Cluster Escape Velocity Profiles
• Simulation-based marginal likelihood for cluster strong lensing cosmology
• Discrepancy-based Inference for Intractable Generative Models using Quasi-Monte Carlo
• Layered Synthesis of Latent Gaussian Trees
• An effective likelihood-free approximate computing method with statistical inferential guarantees
• Parameter Estimation for RANS Models Using Approximate Bayesian Computation
• Deep inference of simulated strong lenses in ground-based surveys
• A review of Approximate Bayesian Computation methods via density estimation: inference for simulator-models
• IQ Collaboratory III: The Empirical Dust Attenuation Framework -- Taking Hydrodynamical Simulations with a Grain of Dust