Normalizing flows for random fields in cosmology

Rouhiainen, Adam; Giri, Utkarsh; Münchmeyer, Moritz

doi:10.48550/arxiv.2105.12024

Cited by 5 publications

(4 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, in order to have a model which is completely physical and can also be used to produce filamentary structures, non-Gaussian features should be introduced. This is a nontrivial task, but previous works in which the nonlinear large-scale structure of the Universe is predicted by deep neural networks using the Zel'dovich approximation (He et al 2019) or even using normalizing flows (Rouhiainen et al 2021) could be used. Another direction of future improvement could be that of the refinement of the algorithm defining our simulator, by finding and easing its bottlenecks.…”

Section: Discussionmentioning

confidence: 99%

Towards reconstructing the halo clustering and halo mass function of N-body simulations using neural ratio estimation

Dimitriou¹,

Weniger²,

Correa³

2022

Preprint

View full text Add to dashboard Cite

High-resolution cosmological N-body simulations are excellent tools for modelling the formation and clustering of dark matter haloes. These simulations suggest complex physical theories of halo formation governed by a set of effective physical parameters. Our goal is to extract these parameters and their uncertainties in a Bayesian context. We make a step towards automatising this process by directly comparing dark matter density projection maps extracted from cosmological simulations, with density projections generated from an analytical halo model. The model is based on a toy implementation of two body correlation functions. To accomplish this we use marginal neural ratio estimation, an algorithm for simulation-based inference that allows marginal posteriors to be estimated by approximating marginal likelihood-to-evidence ratios with a neural network. In this case, we train a neural network with mock images to identify the correct values of the physical parameters that produced a given image. Using the trained neural network on cosmological N-body simulation images we are able to reconstruct the halo mass function, to generate mock images similar to the N-body simulation images and to identify the lowest mass of the haloes of those images, provided that they have the same clustering with our training data. Our results indicate that this is a promising approach in the path towards developing cosmological simulations assisted by neural networks.

show abstract

Section: Discussionmentioning

confidence: 99%

Towards reconstructing the halo clustering and halo mass function of N-body simulations using neural ratio estimation

Dimitriou¹,

Weniger²,

Correa³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…There, Neural Networks are used to learn a mapping between sets of model parameters and their corresponding simulated data, so that they can automatically extract features, marginalise over nuisance parameters, learn a likelihood function, or ultimately produce a posterior distribution of the model parameters when fed real experimental data. Recent development and applications in Cosmology and Astrophysics can be found in [43][44][45][46][47][48][49][50][51][52][53][54]. The claimed advantages are that they may discover or take into account features in the data that are not captured by summary statistics or observables, and the lack of need to formulate a likelihood, which can be complex or prohibitively expensive in some cases.…”

Section: Introductionmentioning

confidence: 99%

Fast and robust Bayesian inference using Gaussian processes with GPry

El Gammal,

Schöneberg,

Torrado

et al. 2023

J. Cosmol. Astropart. Phys.

View full text Add to dashboard Cite

We present the GPry algorithm for fast Bayesian inference of general (non-Gaussian) posteriors with a moderate number of parameters. GPry does not need any pre-training, special hardware such as GPUs, and is intended as a drop-in replacement for traditional Monte Carlo methods for Bayesian inference. Our algorithm is based on generating a Gaussian Process surrogate model of the log-posterior, aided by a Support Vector Machine classifier that excludes extreme or non-finite values. An active learning scheme allows us to reduce the number of required posterior evaluations by two orders of magnitude compared to traditional Monte Carlo inference. Our algorithm allows for parallel evaluations of the posterior at optimal locations, further reducing wall-clock times. We significantly improve performance using properties of the posterior in our active learning scheme and for the definition of the GP prior. In particular we account for the expected dynamical range of the posterior in different dimensionalities. We test our model against a number of synthetic and cosmological examples. GPry outperforms traditional Monte Carlo methods when the evaluation time of the likelihood (or the calculation of theoretical observables) is of the order of seconds; for evaluation times of over a minute it can perform inference in days that would take months using traditional methods. GPry is distributed as an open source Python package (pip install gpry) and can also be found at https://github.com/jonaselgammal/GPry.

show abstract

“…However, the advantages of using NF over other methods is the ability to learn the exact likelihood function to perform either inference or generate new diverse examples by inverting the flow transformations. NF methods have been very successful in generating random cosmological fields (Rouhiainen et al 2021), simulating galaxy images (Lanusse et al 2021), performing likelihood-free inference (e.g., Alsing et al 2019), and modeling color-magnitude diagrams (Cranmer et al 2019). NF attempts to learn the mapping between a standard Gaussian field and the more complex density distribution of the observable (in this case the HI maps).…”

Section: Introductionmentioning

confidence: 99%

HIFlow: Generating Diverse Hi Maps and Inferring Cosmology while Marginalizing over Astrophysics Using Normalizing Flows

Hassan

Villaescusa-Navarro

Wandelt

et al. 2022

ApJ

View full text Add to dashboard Cite

A wealth of cosmological and astrophysical information is expected from many ongoing and upcoming large-scale surveys. It is crucial to prepare for these surveys now and develop tools that can efficiently extract most information. We present HIFlow: a fast generative model of the neutral hydrogen (Hi) maps that is conditioned only on cosmology (Ω m and σ 8) and designed using a class of normalizing flow models, the masked autoregressive flow. HIFlow is trained on the state-of-the-art simulations from the Cosmology and Astrophysics with MachinE Learning Simulations (CAMELS) project. HIFlow has the ability to generate realistic diverse maps without explicitly incorporating the expected two-dimensional maps structure into the flow as an inductive bias. We find that HIFlow is able to reproduce the CAMELS average and standard deviation Hi power spectrum within a factor of ≲2, scoring a very high R 2 > 90%. By inverting the flow, HIFlow provides a tractable high-dimensional likelihood for efficient parameter inference. We show that the conditional HIFlow on cosmology is successfully able to marginalize over astrophysics at the field level, regardless of the stellar and AGN feedback strengths. This new tool represents a first step toward a more powerful parameter inference, maximizing the scientific return of future Hi surveys, and opening a new avenue to minimize the loss of complex information due to data compression down to summary statistics.

show abstract

Normalizing flows for random fields in cosmology

Cited by 5 publications

References 19 publications

Towards reconstructing the halo clustering and halo mass function of N-body simulations using neural ratio estimation

Towards reconstructing the halo clustering and halo mass function of N-body simulations using neural ratio estimation

Fast and robust Bayesian inference using Gaussian processes with GPry

HIFlow: Generating Diverse Hi Maps and Inferring Cosmology while Marginalizing over Astrophysics Using Normalizing Flows

Contact Info

Product

Resources

About