Cosmological parameter estimation via iterative emulation of likelihoods

@article{PellejeroIbaez2019CosmologicalPE,
  title={Cosmological parameter estimation via iterative emulation of likelihoods},
  author={Marcos Pellejero-Iba{\~n}ez and Raul E Angulo and Giovanni Aric{\'o} and Matteo Zennaro and Sergio Contreras and Jens St{\"u}cker},
  journal={arXiv: Cosmology and Nongalactic Astrophysics},
  year={2019}
}
The interpretation of cosmological observables requires the use of increasingly sophisticated theoretical models. Since these models are becoming computationally very expensive and display non-trivial uncertainties, the use of standard Bayesian algorithms for cosmological inferences, such as MCMC, might become inadequate. Here, we propose a new approach to parameter estimation based on an iterative Gaussian emulation of the target likelihood function. This requires a minimal number of… Expand

Figures and Tables from this paper

3 per cent-accurate predictions for the clustering of dark matter, haloes, and subhaloes, over a wide range of cosmologies and scales
Predicting the spatial distribution of objects as a function of cosmology is an essential ingredient for the exploitation of future galaxy surveys. In this paper we show that a specially-designedExpand
Cosmological Parameter Estimation and Inference using Deep Summaries
The ability to obtain reliable point estimates of model parameters is of crucial importance in many fields of physics. This is often a difficult task given that the observed data can have a very highExpand
The BACCO simulation project: a baryonification emulator with neural networks
We present a neural network emulator for baryonic effects in the non-linear matter power spectrum. We calibrate this emulator using more than 50 000 measurements in a 15D parameter space, varyingExpand
The BACCO simulation project: exploiting the full power of large-scale structure for cosmology
We present the BACCO project, a simulation framework specially designed to provide highly-accurate predictions for the distribution of mass, galaxies, and gas as a function of cosmologicalExpand
Boosting Monte Carlo sampling with a non-Gaussian fit
We propose a new method, called MonteCarlo Posterior Fit, to boost the MonteCarlo sampling of likelihood (posterior) functions. The idea is to approximate the posterior function by an analyticalExpand
Simultaneous modelling of matter power spectrum and bispectrum in the presence of baryons
We demonstrate that baryonification algorithms, which displace particles in gravity-only simulations according to physically-motivated prescriptions, can simultaneously capture the impact of baryonicExpand

References

SHOWING 1-10 OF 87 REFERENCES
Reconstructing probability distributions with Gaussian processes
Modern cosmological analyses constrain physical parameters using Markov Chain Monte Carlo (MCMC) or similar sampling techniques. Oftentimes, these techniques are computationally expensive to runExpand
Bayesian Optimization for Likelihood-Free Inference of Simulator-Based Statistical Models
TLDR
This paper proposes a strategy which combines probabilistic modeling of the discrepancy with optimization to facilitate likelihood-free inference and is shown to accelerate the inference through a reduction in the number of required simulations by several orders of magnitude. Expand
Fast $\epsilon$-free Inference of Simulation Models with Bayesian Conditional Density Estimation
TLDR
A new approach to likelihood-free inference based on Bayesian conditional density estimation is proposed, which requires fewer model simulations than Monte Carlo ABC methods need to produce a single sample from an approximate posterior. Expand
Approximate Bayesian computation via regression density estimation
TLDR
The contribution of the present paper is to consider regression density estimation techniques to approximate the likelihood in the ABC setting, which builds on recently developed marginal adaptation density estimators by extending them for conditional density estimation. Expand
Efficient Acquisition Rules for Model-Based Approximate Bayesian Computation
TLDR
This paper proposes to compute the uncertainty in the ABC posterior density, which is due to a lack of simulations to estimate this quantity accurately, and defines a loss function that measures this uncertainty and proposes to select the next evaluation location to minimise the expected loss. Expand
Modelling the large-scale mass density field of the universe as a function of cosmology and baryonic physics
We present and test a framework that models the three-dimensional distribution of mass in the Universe as a function of cosmological and astrophysical parameters. Our approach combines two differentExpand
On Sparse Variational Methods and the Kullback-Leibler Divergence between Stochastic Processes
TLDR
A substantial generalization of the literature on variational framework for learning inducing variables is given and a new proof of the result for infinite index sets is given which allows inducing points that are not data points and likelihoods that depend on all function values. Expand
GPy: A Gaussian process framework in python
  • PASP,
  • 2013
The Cosmic Linear Anisotropy Solving System (CLASS). Part II: Approximation schemes
Boltzmann codes are used extensively by several groups for constraining cosmo- logical parameters with Cosmic Microwave Background and Large Scale Structure data. This activity is computationallyExpand
Higher order Hamiltonian Monte Carlo sampling for cosmological large-scale structure analysis
We investigate higher order symplectic integration strategies within Bayesian cosmic density field reconstruction methods. In particular, we study the fourth-order discretization of HamiltonianExpand
...
1
2
3
4
5
...