Cosmic Kite: Auto-encoding the Cosmic Microwave Background

Abstract

In this work we present the results of the study of the cosmic microwave background TT power spectrum through auto-encoders in which the latent variables are the cosmological parameters. This method was trained and calibrated using a data-set composed by 80000 power spectra from random cosmologies computed numerically with the CAMB code. Due to the specific architecture of the auto-encoder, the encoder part is a model that estimates the maximum-likelihood parameters from a given power spectrum. On the other hand, the decoder part is a model that computes the power spectrum from the cosmological parameters and can be used as a forward model in a fully Bayesian analysis. We show that the encoder is able to estimate the true cosmological parameters with a precision varying from $\approx 0.004\%$ to $\approx 0.2\%$ (depending on the cosmological parameter), while the decoder computes the power spectra with a mean percentage error of $\approx 0.0018\%$ for all the multipole range. We also demonstrate that the decoder recovers the expected trends when varying the cosmological parameters one by one, and that it does not introduce any significant bias on the estimation of cosmological parameters through a Bayesian analysis. These studies gave place to the Cosmic Kite python software that is publicly available and can be downloaded and installed from https://github.com/Martindelosrios/cosmic-kite. Although this algorithm does not improve the precision of the measurements compared with the traditional methods, it reduces significantly the computation time and represents the first attempt towards forcing the latent variables to have a physical interpretation.

arXiv commentAccepted for its publication in MNRAS

References in corpus (16)

• HEALPix -- a Framework for High Resolution Discretization, and Fast Analysis of Data Distributed on the Sphere
• Efficient Computation of CMB anisotropies in closed FRW models
• The Cosmic Linear Anisotropy Solving System (CLASS) II: Approximation schemes
• A 2.4% Determination of the Local Value of the Hubble Constant
• Planck 2018 results. V. CMB power spectra and likelihoods
• MASTER of the CMB Anisotropy Power Spectrum: A Fast Method for Statistical Analysis of Large and Complex CMB Data Sets
• A unified pseudo-$C_{\ell }$ framework
• COSMOPOWER: emulating cosmological power spectra for accelerated Bayesian inference from next-generation surveys
• DeepSphere: Efficient spherical Convolutional Neural Network with HEALPix sampling for cosmological applications
• Secondary anisotropies of the CMB
• Fast cosmological parameter estimation using neural networks
• {\sc CosmoNet}: fast cosmological parameter estimation in non-flat models using neural networks
• A practical guide to Basic Statistical Techniques for Data Analysis in Cosmology
• Inpainting CMB maps using Partial Convolutional Neural Networks
• Data driven foreground clustering approach to component separation in multifrequency CMB experiments: A new Planck CMB map
• Foreground removal from WMAP 7yr polarization maps using an MLP neural network

Cited by in corpus (3)

• Neural Networks for cosmological model selection and feature importance using Cosmic Microwave Background data
• Conditional variational autoencoders for cosmological model discrimination and anomaly detection in cosmic microwave background power spectra
• Dark Matter profiles of "in silico" galaxies: deep learning inference