This github attempts to maintain a comprehensive list of published machine learning applications to cosmology, organized by subject matter and arxiv posting date. Each entry contains the paper title, a simple summary of the machine learning methods used in the work, and the arxiv link. If I have missed any cosmology papers that you believe should be included please email me at [email protected] or issue a pull request.

Feel free to cite in any works DOI

I am currently a postdoctoral researcher at the Berkeley Center for Cosmological Physics, broadly working on problems in computational cosmology, but with a great interest in machine learning methods, and just made this for fun and to help anyone with similar interests. Cheers to whoever can find which of the papers below have me as an author

Table of Contents

Section List


A dictionary of all abbreviations for machine learning methods used in this compilation. In general I adopted those used by the authors, except in a few cases. The links are to explanatory articles that I personally like.


Large-Scale Structure

The Large-Scale Structure of the universe is a field that relies on state-of-the art cosmological simulations to address a number of questions. Due to the computational complexity of these simulations, some investigations will remain computationally-infeasible for the forseeable future, and machine learning techniques can have a number of important uses.

Structure Formation

Title ML technique(s) used arxiv link
A First Look at creating mock catalogs with machine learning techniques SVM, kNN
Machine Learning Etudes in Astrophysics: Selection Functions for Mock Cluster Catalogs SVM, GMM
PkANN I&2. Non-linear matter power spectrum interpolation through artificial neural networks NN,
Machine learning and cosmological simulations I.&II. kNN, DT, RF, EXT
Estimating Cosmological Parameters from the Dark Matter Distribution CNN
Painting galaxies into dark matter haloes using machine learning SVR, kNN, MLP, DT, RF, EXT, AdR
Modeling the Impact of Baryons on Subhalo Populations with Machine Learning RF
Fast cosmic web simulations with generative adversarial networks GAN
Machine learning cosmological structure formation RF
A Machine Learning Approach to Galaxy-LSS Classification I: Imprints on Halo Merger Trees SVM
Classifying the Large Scale Structure of the Universe with Deep Neural Networks V-Net
Cosmological Reconstruction From Galaxy Light: Neural Network Based Light-Matter Connection NN
A volumetric deep Convolutional Neural Network for simulation of mock dark matter halo catalogues V-Net
Learning to Predict the Cosmological Structure Formation V-Net
deepCool: Fast and Accurate Estimation of Cooling Rates in Irradiated Gas with Artificial Neural Networks NN, RF, kNN
From Dark Matter to Galaxies with Convolutional Networks V-Net
Painting halos from 3D dark matter fields using Wasserstein mapping networks GAN
Painting with baryons: augmenting N-body simulations with gas using deep generative models GAN, VAE
HIGAN: Cosmic Neutral Hydrogen with Generative Adversarial Networks GAN
A deep learning model to emulate simulations of cosmic reionization CNN
An interpretable machine learning framework for dark matter halo formation BDT
Cosmological N-body simulations: a challenge for scalable generative models GAN
Cosmological parameter estimation from large-scale structure deep learning CNN
Neural physical engines for inferring the halo mass distribution function NPE
A Hybrid Deep Learning Approach to Cosmological Constraints From Galaxy Redshift Surveys CNN
A black box for dark sector physics: Predicting dark matter annihilation feedback with conditional GANs cGAN
Learning neutrino effects in Cosmology with Convolutional Neural Networks V-Net
Predicting dark matter halo formation in N-body simulations with deep regression networks V-Net
Probabilistic cosmic web classification using fast-generated training data RF
Super-resolution emulator of cosmological simulations using deep physical models WGAN
Baryon acoustic oscillations reconstruction using convolutional neural networks CNN
Emulation of cosmological mass maps with conditional generative adversarial networks GAN
Towards Universal Cosmological Emulators with Generative Adversarial Networks GAN
Nonlinear 3D Cosmic Web Simulation with Heavy-Tailed Generative Adversarial Networks GAN
GalaxyNet: Connecting galaxies and dark matter haloes with deep neural networks and reinforcement learning in large volumes RF, NN
Discovering Symbolic Models from Deep Learning with Inductive Biases GNN
Teaching neural networks to generate Fast Sunyaev Zel’dovich Maps V-Net
HInet: Generating neutral hydrogen from dark matter with neural networks CNN
Machine Learning the Fates of Dark Matter Subhalos: A Fuzzy Crystal Ball RF, BDT
Learning effective physical laws for generating cosmological hydrodynamics with Lagrangian Deep Learning LDL
AI-assisted super-resolution cosmological simulations GAN
Encoding large scale cosmological structure with Generative Adversarial Networks GAN
Deep learning insights into cosmological structure formation CNN
SHAPing the Gas: Understanding Gas Shapes in Dark Matter Haloes with Interpretable Machine Learning XGBoost
dm2gal: Mapping Dark Matter to Galaxies with Neural Networks CNN
Fast and Accurate Non-Linear Predictions of Universes with Deep Learning V-Net
The BACCO Simulation Project: A baryonification emulator with Neural Networks NN
dm2gal: Mapping Dark Matter to Galaxies with Neural Networks CNN
Fast and Accurate Non-Linear Predictions of Universes with Deep Learning V-Net
Identifying Cosmological Information in a Deep Neural Network CNN
CosmicRIM : Reconstructing Early Universe by Combining Differentiable Simulations with Recurrent Inference Machines RIM
AI-assisted super-resolution cosmological simulations II: Halo substructures, velocities and higher order statistics GAN
Cosmic Velocity Field Reconstruction Using AI V-Net
Normalizing flows for random fields in cosmology NF
Classification algorithms applied to structure formation simulations RF
Fast, high-fidelity Lyman α forests with convolutional neural networks V-Net
HyPhy: Deep Generative Conditional Posterior Mapping of Hydrodynamical Physics VAE
Predicting halo occupation and galaxy assembly bias with machine learning RF
Finding universal relations in subhalo properties with artificial intelligence NN
Multifield Cosmology with Artificial Intelligence CNN
Robust marginalization of baryonic effects for cosmological inference at the field level CNN

Structure Identification

Title ML technique(s) used arxiv link
A Machine Learning Approach for Dynamical Mass Measurements of Galaxy Clusters SDM,
A Deep Learning Approach to Galaxy Cluster X-ray Masses CNN
An application of machine learning techniques to galaxy cluster mass estimation using the MACSIS simulations OLR, RR, BRR, KRR, SVR, DT, BDT, ADA, kNN
Prediction of galaxy halo masses in SDSS DR7 via a machine learning approach XGBoost, RF, NN
A Robust and Efficient Deep Learning Method for Dynamical Mass Measurements of Galaxy Clusters CNN
Multiwavelength cluster mass estimates and machine learning GB, RF
Self-supervised Learning with Physics-aware Neural Networks I: Galaxy Model Fitting AE
Using X-Ray Morphological Parameters to Strengthen Galaxy Cluster Mass Estimates via Machine Learning RF
Large-scale structures in the LCDM Universe: network analysis and machine learning XGBoost
Dynamical mass inference of galaxy clusters with neural flows NF (MADE)
Mass Estimation of Galaxy Clusters with Deep Learning I: Sunyaev-Zel’dovich Effect U-Net
Galaxy cluster mass estimation with deep learning and hydrodynamical simulations CNN
Mass Estimation of Galaxy Clusters with Deep Learning II: CMB Cluster Lensing U-NET
Multiwavelength classification of X-ray selected galaxy cluster candidates using convolutional neural networks CNN
Anomaly detection in Astrophysics: a comparison between unsupervised Deep and Machine Learning on KiDS data AE, RF
Approximate Bayesian Uncertainties on Deep Learning Dynamical Mass Estimates of Galaxy Clusters BNN
A deep learning view of the census of galaxy clusters in IllustrisTNG CNN
Revealing the Local Cosmic Web by Deep Learning V-Net
Simulation-based inference of dynamical galaxy cluster masses with 3D convolutional neural networks CNN
Weak-lensing Mass Reconstruction of Galaxy Clusters with Convolutional Neural Network CNN
DeepSZ: Identification of Sunyaev-Zel’dovich Galaxy Clusters using Deep Learning CNN
DeepMerge II: Building Robust Deep Learning Algorithms for Merging Galaxy Identification Across Domains CNN


Reionization and 21cm

In cosmology, the process of Reionization refers to the period when our universe went from the “Dark Ages” before major star and galaxy formation, to the ionized state we see today.

Title ML technique(s) used arxiv link
A machine-learning approach to measuring the escape of ionizing radiation from galaxies in the reionization epoch LR
Analysing the 21 cm signal from the epoch of reionization with artificial neural networks NN
Emulation of reionization simulations for Bayesian inference of astrophysics parameters using neural networks NN
Reionization Models Classifier using 21cm Map Deep Learning CNN
Deep learning from 21-cm images of the Cosmic Dawn CNN
Identifying Reionization Sources from 21cm Maps using Convolutional Neural Networks CNN
Evaluating machine learning techniques for predicting power spectra from reionization simulations SVM, MLP, GPR
Improved supervised learning methods for EoR parameters reconstruction CNN
Constraining the astrophysics and cosmology from 21cm tomography using deep learning with the SKA CNN
Emulating the Global 21-cm Signal from Cosmic Dawn and Reionization NN
21cm Global Signal Extraction: Extracting the 21cm Global Signal using Artificial Neural Networks NN
A unified framework for 21cm tomography sample generation and parameter inference with Progressively Growing GANs GAN
Beyond the power spectrum – I: recovering H II bubble size distribution from 21 cm power spectrum with artificial neural networks NN
Foreground modelling via Gaussian process regression: an application to HERA data GP
Predicting 21cm-line map from Lyman α emitter distribution with Generative Adversarial Networks GAN
Constraining the Reionization History using Bayesian Normalizing Flows NF
Deep-Learning Study of the 21cm Differential Brightness Temperature During the Epoch of Reionization CNN
Removing Astrophysics in 21 cm maps with Neural Networks CNN
Deep Forest: Neural Network reconstruction of the Lyman-alpha forest NN
deep21: a Deep Learning Method for 21cm Foreground Removal U-Net
Analysing the Epoch of Reionization with three-point correlation functions and machine learning techniques NN
Using Artificial Neural Networks to extract the 21-cm Global Signal from the EDGES data NN
Modeling assembly bias with machine learning and symbolic regression RF, SR
Reconstructing Patchy Reionization with Deep Learning U-Net
Deep learning approach for identification of HII regions during reionization in 21-cm observations U-Net
GLOBALEMU: A novel and robust approach for emulating the sky-averaged 21-cm signal from the cosmic dawn and epoch of reionisation NN
Machine learning galaxy properties from 21 cm lightcones: impact of network architectures and signal contamination CNN
21cmVAE: A VAE-based Emulator of the 21-cm Global Signal VAE
Probing Ultra-light Axion Dark Matter from 21cm Tomography using Convolutional Neural Networks CNN
Deep Forest: Neural Network reconstruction of the Lyman-alpha forest NN


Gravitational Lensing

Gravitational lensing in cosmology refers to the bending of light due to mass between the source and Earth. This effect is very useful for inferring properties of the total mass distribution in our Universe, which is dominated by dark matter that we cannot see electromagnetically. Gravitational lensing comes in two types: weak and strong.

Strong gravitational lensing refers to the cases where the lensing effect (e.g. multiple images, clear shape distortions) is strong enough to be seen by the human eye, or equivalent, on an astronomical image. This only happens when a massive galaxy cluster lies between us and some background galaxies

Weak gravitational lensing refers to the global effect that almost all far away galaxies are gravitationally lensed by a small amount, which changes their observed shape by roughly 1%. This can only be measured statistically when given a large number of samples, and not on an object-to-object basis.

Weak Lensing

Title ML technique(s) used arxiv link
Bias-Free Shear Estimation using Artificial Neural Networks NN
Hopfield Neural Network deconvolution for weak lensing measurement HNN
CosmoGAN: creating high-fidelity weak lensing convergence maps using Generative Adversarial Networks GAN
Cosmological model discrimination with Deep Learning CNN
Non-Gaussian information from weak lensing data via deep learning CNN
Learning from deep learning: better cosmological parameter inference from weak lensing maps CNN
Weak-lensing shear measurement with machine learning: teaching artificial neural networks about feature noise NN
Cosmological constraints from noisy convergence maps through deep learning CNN
Weak lensing shear estimation beyond the shape-noise limit: a machine learning approach CNN
On the dissection of degenerate cosmologies with machine learning CNN
Distinguishing standard and modified gravity cosmologies with machine learning CNN
Denoising Weak Lensing Mass Maps with Deep Learning GAN
Weak lensing cosmology with convolutional neural networks on noisy data CNN
Galaxy shape measurement with convolutional neural networks CNN
Cosmological constraints with deep learning from KiDS-450 weak lensing maps CNN
Deep learning dark matter map reconstructions from DES SV weak lensing data U-Net
Decoding Cosmological Information in Weak-Lensing Mass Maps with Generative Adversarial Networks GAN
Parameter Inference for Weak Lensing using Gaussian Processes and MOPED GP
Shear measurement bias II: a fast machine learning calibration method NN
Interpreting deep learning models for weak lensing CNN
Shear measurement bias II: a fast machine learning calibration method MLP
Probabilistic Mapping of Dark Matter by Neural Score Matching DE
Higher order statistics of shear field: a machine learning approach kNN, SVM, GP, RF, etc..
Simultaneously constraining cosmology and baryonic physics via deep learning from weak lensing CNN

Strong Lensing

Title ML technique(s) used arxiv link
A neural network gravitational arc finder based on the Mediatrix filamentation method NN
CMU DeepLens: deep learning for automatic image-based galaxy-galaxy strong lens finding CNN
Automated Lensing Learner: Automated Strong Lensing Identification with a Computer Vision Technique HoG
Finding strong lenses in CFHTLS using convolutional neural networks CNN
Fast automated analysis of strong gravitational lenses with convolutional neural networks CNN
Uncertainties in Parameters Estimated with Neural Networks: Application to Strong Gravitational Lensing NN
The Strong Gravitational Lens Finding Challenge SVM, CNN
Testing convolutional neural networks for finding strong gravitational lenses in KiDS CNN
Analyzing interferometric observations of strong gravitational lenses with recurrent and convolutional neural networks RNN, CNN
Data-Driven Reconstruction of Gravitationally Lensed Galaxies using Recurrent Inference Machines RIM, CNN
Finding Strong Gravitational Lenses in the DESI DECam Legacy Survey CNN
Mining for Dark Matter Substructure: Inferring subhalo population properties from strong lenses with machine learning NN
Deep Learning the Morphology of Dark Matter Substructure CNN
Circumventing Lens Modeling to Detect Dark Matter Substructure in Strong Lens Images with Convolutional Neural Networks CNN
Differentiable Strong Lensing: Uniting Gravity and Neural Nets through Differentiable Probabilistic Programming VAE
Identifying Strong Lenses with Unsupervised Machine Learning using Convolutional Autoencoder VAE
HOLISMOKES II. Identifying galaxy-scale strong gravitational lenses in Pan-STARRS using convolutional neural networks CNN
Discovering New Strong Gravitational Lenses in the DESI Legacy Imaging Surveys CNN
Dark Matter Subhalos, Strong Lensing and Machine Learning CNN
Deep Learning for Strong Lensing Search: Tests of the Convolutional Neural Networks and New Candidates from KiDS DR3 CNN
Decoding Dark Matter Substructure without Supervision AE, VAE, AAE
Extracting the Subhalo Mass Function from Strong Lens Images with Image Segmentation U-Net
Detecting Subhalos in Strong Gravitational Lens Images with Image Segmentation U-Net
Hunting for Dark Matter Subhalos in Strong Gravitational Lensing with Neural Networks CNN
Targeted Likelihood-Free Inference of Dark Matter Substructure in Strongly-Lensed Galaxies GP, …
Large-Scale Gravitational Lens Modeling with Bayesian Neural Networks for Accurate and Precise Inference of the Hubble Constant BNN
Strong lens systems search in the Dark Energy Survey using Convolutional Neural Networks CNN
Finding quadruply imaged quasars with machine learning. I. Methods CNN, VAE


Cosmic Microwave Background

The Cosmic Microwave Background (CMB) is the light left over from the period of recombination in the very early Universe, 380,000 years after the beginning. CMB observations are sometimes referred to as “baby pictures of our Universe”, as this light has been travelling for 13.5 billion years just to reach us.

Title ML technique(s) used arxiv link
DeepCMB: Lensing Reconstruction of the Cosmic Microwave Background with Deep Neural Networks CNN
Fast Wiener filtering of CMB maps with Neural Networks U-Net
CMB-GAN: Fast Simulations of Cosmic Microwave background anisotropy maps using Deep Learning GAN
CosmoVAE: Variational Autoencoder for CMB Image Inpainting VAE
Inpainting Galactic Foreground Intensity and Polarization maps using Convolutional Neural Network GAN
Inpainting via Generative Adversarial Networks for CMB data analysis GAN
Full-sky Cosmic Microwave Background Foreground Cleaning Using Machine Learning BNN
Foreground model recognition through Neural Networks for CMB B-mode observations NN
Inpainting CMB maps using Partial Convolutional Neural Networks U-Net
ForSE: a GAN based algorithm for extending CMB foreground models to sub-degree angular scales GAN
A Generative Model of Galactic Dust Emission Using Variational Inference VAE
A convolutional-neural-network estimator of CMB constraints on dark matter energy injection CNN
An Unbiased Estimator of the Full-sky CMB Angular Power Spectrum using Neural Networks NN
MillimeterDL: Deep Learning Simulations of the Microwave Sky U-Net
Reconstructing Cosmic Polarization Rotation with ResUNet-CMB U-Net



This section has a variety of machine learning papers used for various observational applications.


This section is definitely not exhaustive – there is a massive amount of work in this subject area.

Title ML technique(s) used arxiv link
ANNz: estimating photometric redshifts using artificial neural networks NN
Estimating Photometric Redshifts Using Support Vector Machines SVM
Robust Machine Learning Applied to Astronomical Data Sets. II. Quantifying Photometric Redshifts for Quasars Using Instance-based Learning kNN
Robust Machine Learning Applied to Astronomical Data Sets. III. Probabilistic Photometric Redshifts for Galaxies and Quasars in the SDSS and GALEX kNN
ArborZ: Photometric Redshifts Using Boosted Decision Trees BDT
Unsupervised self-organised mapping: a versatile empirical tool for object selection, classification and redshift estimation in large surveys SOM
Can Self-Organizing Maps accurately predict photometric redshifts? SOM
TPZ : Photometric redshift PDFs and ancillary information by using prediction trees and random forests RF
Estimating Photometric Redshifts of Quasars via K-nearest Neighbor Approach Based on Large Survey Databases kNN
An approach to the analysis of SDSS spectroscopic outliers based on Self-Organizing Maps SOM
Using neural networks to estimate redshift distributions. An application to CFHTLenS NN
SOMz: photometric redshift PDFs with self organizing maps and random atlas SOM
Feature importance for machine learning redshifts applied to SDSS galaxies NN, ADA
GAz: A Genetic Algorithm for Photometric Redshift Estimation GA
Anomaly detection for machine learning redshifts applied to SDSS galaxies ADA, SOM, BDT
Measuring photometric redshifts using galaxy images and Deep Neural Networks CNN, ADA
A Sparse Gaussian Process Framework for Photometric Redshift Estimation NN, GPR
ANNz2 – photometric redshift and probability distribution function estimation using machine learning NN, BDT
DNF – Galaxy photometric redshift by Directional Neighbourhood Fitting kNN
Photometric Redshift Estimation for Quasars by Integration of KNN and SVM kNN, SVM
Stacking for machine learning redshifts applied to SDSS galaxies SOM, DT
GPz: Non-stationary sparse Gaussian processes for heteroscedastic uncertainty estimation in photometric redshifts GPR
Photo-z with CuBANz: An improved photometric redshift estimator using Clustering aided Back Propagation Neural network NN
Photometric redshift estimation via deep learning. Generalized and pre-classification-less, image based, fully probabilistic redshifts RF, MDN, DCMDN
Photometric redshifts for the Kilo-Degree Survey. Machine-learning analysis with artificial neural networks NN, BDT
Estimating Photometric Redshifts for X-ray sources in the X-ATLAS field, using machine-learning techniques RF
Deep learning approach for classifying, detecting and predicting photometric redshifts of quasars in the Sloan Digital Sky Survey stripe 82 CNN, kNN, SVM, RF, GPR
Return of the features. Efficient feature selection and interpretation for photometric redshifts kNN
Photometric redshifts from SDSS images using a Convolutional Neural Network CNN
Estimating redshift distributions using Hierarchical Logistic Gaussian processes GPR
Gaussian Mixture Models for Blended Photometric Redshifts GMM
Photometric Redshift Calibration with Self Organising Maps SOM
PS1-STRM: Neural network source classification and photometric redshift catalogue for PS1 NN
Reliable Photometric Membership (RPM) of Galaxies in Clusters. I. A Machine Learning Method and its Performance in the Local Universe SVM
PhotoWeb redshift: boosting photometric redshift accuracy with large spectroscopic surveys CNN
The PAU Survey: Photometric redshifts using transfer learning from simulations MDN
KiDS+VIKING-450: Improved cosmological parameter constraints from redshift calibration with self-organising maps SOM
Determining the systemic redshift of Lyman-α emitters with neural networks and improving the measured large-scale clustering NN
Photometric selection and redshifts for quasars in the Kilo-Degree Survey Data Release 4 RF, XGBoost, NN
Photometric Redshift Estimation with a Convolutional Neural Network: NetZ CNN
A machine learning approach to galaxy properties: joint redshift-stellar mass probability distributions with Random Forest RF
Spectroscopic and Photometric Redshift Estimation by Neural Networks For the China Space Station Optical Survey (CSS-OS) NN
Estimating Galactic Distances From Images Using Self-supervised Representation Learning SSL
QSO photometric redshifts using machine learning and neural networks kNN, DT, NN
Benchmarking and Scalability of Machine Learning Methods for Photometric Redshift Estimation RF, BDT, kNN
Z-Sequence: Photometric redshift predictions for galaxy clusters with sequential random k-nearest neighbours kNN
Probabilistic photo-z machine learning models for X-ray sky surveys RF
Non-Sequential Neural Network for Simultaneous, Consistent Classification and Photometric Redshifts of OTELO Galaxies NN
Using a Neural Network Classifier to Select Galaxies with the Most Accurate Photometric Redshifts NN
Investigating Deep Learning Methods for Obtaining Photometric Redshift Estimations from Images RF, CNN

Other Observational

Title ML technique(s) used arxiv link
Use of neural networks for the identification of new z>=3.6 QSOs from FIRST-SDSS DR5 NN
Estimating the Mass of the Local Group using Machine Learning Applied to Numerical Simulations NN
A probabilistic approach to emission-line galaxy classification GMM
Deep Learning of Quasar Spectra to Discover and Characterize Damped Lya Systems CNN
An automatic taxonomy of galaxy morphology using unsupervised machine learning SOM
Learning from the machine: interpreting machine learning algorithms for point- and extended- source classification RF, ADA, EXT, BDT, MINT, TINT
Predicting the Neutral Hydrogen Content of Galaxies From Optical Data Using Machine Learning OLR, RF, BDT, kNN, SVM, NN
Star-galaxy classification in the Dark Energy Survey Y1 dataset SVM, ADA
Classifying galaxy spectra at 0.5<z<1 with self-organizing maps SOM
Knowledge transfer of Deep Learning for galaxy morphology from one survey to another CNN
Classification of Broad Absorption Line Quasars with a Convolutional Neural Network CNN
Generative deep fields: arbitrarily sized, random synthetic astronomical images through deep learning GAN
Deconfusing intensity maps with neural networks CNN
Deep-CEE I: Fishing for Galaxy Clusters with Deep Neural Nets RCNN
Improving Galaxy Clustering Measurements with Deep Learning: analysis of the DECaLS DR7 data NN
What can Machine Learning tell us about the background expansion of the Universe? GA
A deep learning approach to cosmological dark energy models BNN+RNN
Reconstructing Functions and Estimating Parameters with Artificial Neural Network: a test with Hubble parameter and SNe Ia NN, RNN, LSTM, GRU
Multi-wavelength properties of radio and machine-learning identified counterparts to submillimeter sources in S2COSMOS SVM, XGBoost
Machine learning computation of distance modulus for local galaxies kNN, BDT, NN
MILCANN : A neural network assessed tSZ map for galaxy cluster detection NN
Machine Learning meets the redshift evolution of the CMB Temperature GA
Inverse Cosmography: testing the effectiveness of cosmographic polynomials using machine learning RNN+BNN
Deblending galaxies with Variational Autoencoders: a joint multi-band, multi-instrument approach VAE
Fully probabilistic quasar continua predictions near Lyman-α with conditional neural spline flows NF
Artificial intelligence and quasar absorption system modelling; application to fundamental constants at high redshift AI
Deep learning the astrometric signature of dark matter substructure CNN
Beyond the Hubble Sequence — Exploring Galaxy Morphology with Unsupervised Machine Learning VAE
Deep Learning for Line Intensity Mapping Observations: Information Extraction from Noisy Maps GAN
Peculiar Velocity Estimation from Kinetic SZ Effect using Deep Neural Networks CNN
Machine learning forecasts of the cosmic distance duality relation with strongly lensed gravitational wave events GP, GA
DeepShadows: Separating Low Surface Brightness Galaxies from Artifacts using Deep Learning CNN
Model independent calibrations of gamma ray bursts using machine learning RF, NN
Self-Supervised Representation Learning for Astronomical Images SSL
An Active Galactic Nucleus Recognition Model based on Deep Neural Network NN
A Machine Learning Approach to Measuring the Quenched Fraction of Low-Mass Satellites Beyond the Local Group NN
The PAU survey: Estimating galaxy photometry with deep learning CNN
Anomaly detection in Hyper Suprime-Cam galaxy images with generative adversarial networks GAN, AE
Euclid preparation: XVI. Forecasts for galaxy morphology with the Euclid Survey using Deep Generative Models VAE
Planck Limits on Cosmic String Tension Using Machine Learning CNN
Morphological classification of compact and extended radio galaxies using convolutional neural networks and data augmentation techniques CNN
Capturing the physics of MaNGA galaxies with self-supervised Machine Learning SSL
Galaxy Deblending using Residual Dense Neural networks RDN


Parameter Estimation

Cosmological parameter estimation is the mechanism of inferring the contents and evolution of our universe from observations. This topic is quite broad, and therefore parameter estimation papers with a focus on an individual experiment/dataset can be found in other sections (e.g. the Reionization and 21cm section). Note this section is unfinished

Title ML technique(s) used arxiv link
Bayesian emulator optimisation for cosmology: application to the Lyman-alpha forest GP
Fast likelihood-free cosmology with neural density estimators and active learning MDN, NF (MAF)
Accelerated Bayesian inference using deep learning NN
Cosmic Inference: Constraining Parameters With Observations and Highly Limited Number of Simulations GP
Euclid-era cosmology for everyone: Neural net assisted MCMC sampling for the joint 3×2 likelihood NN
Parameters Estimation for the Cosmic Microwave Background with Bayesian Neural Networks BNNs
Flow-Based Likelihoods for Non-Gaussian Inference NF
Nearest Neighbor distributions: new statistical measures for cosmological clustering kNN-CDF
Likelihood-free inference with neural compression of DES SV weak lensing map statistics NF
Neural networks as optimal estimators to marginalize over baryonic effects NN
Solving high-dimensional parameter inference: marginal posterior densities & Moment Networks NF
Accelerating MCMC algorithms through Bayesian Deep Networks BNN
Seeking New Physics in Cosmology with Bayesian Neural Networks I: Dark Energy and Modified Gravity BNN
Unsupervised Resource Allocation with Graph Neural Networks GNN
Machine-driven searches for cosmological physics IM



Contained here are some machine learning tools that are specifically designed for the computational challenges of cosmology.

Title ML technique(s) used arxiv link
CosmoFlow: Using Deep Learning to Learn the Universe at Scale CNN
DeepSphere: Efficient spherical Convolutional Neural Network with HEALPix sampling for cosmological applications CNN
Convolutional Neural Networks on the HEALPix sphere: a pixel-based algorithm and its application to CMB data analysis CNN
CosmicNet I: Physics-driven implementation of neural networks within Boltzmann-Einstein solvers NN
FlowPM: Distributed TensorFlow Implementation of the FastPM Cosmological N-body Solver TF
Towards Machine Learning-Based Meta-Studies: Applications to Cosmological Parameters NLP
Equivariant Networks for Pixelized Spheres

Public Datasets

Contained here are some cosmological machine learning datasets.

Title arxiv link github link
Aemulus Project
The Quijote simulations
The CAMELS project: Cosmology and Astrophysics with MachinE Learning Simulations
The CAMELS Multifield Dataset: Learning the Universe’s Fundamental Parameters with Artificial Intelligence



Reviews of machine learning in cosmology, and, more broadly, astronomy.

Title arxiv link
Data Mining and Machine Learning in Astronomy
The Role of Machine Learning in the Next Decade of Cosmology
Machine learning and the physical sciences



Thanks to the following people for bringing additional papers to my attention!

Philippe Berger

Dana Simard

Michelle Ntampaka

Farida Farsian

Celia Escamilla-Rivera

Michaël Defferrard

Farida Farsian

Pranath Reddy

Camille Avestruz

Harry Bevins