Machine learning cosmic inflation

Abstract

We present a machine-learning approach, based on the genetic algorithms (GA), that can be used to reconstruct the inflationary potential directly from cosmological data. We create a pipeline consisting of the GA, a primordial code and a Boltzmann code used to calculate the theoretical predictions, and Cosmic Microwave Background (CMB) data. As a proof of concept, we apply our methodology to the Planck CMB data and explore the functional space of single-field inflationary potentials in a non-parametric, yet analytical way. We show that the algorithm easily improves upon the vanilla model of quadratic inflation and proposes slow-roll potentials better suited to the data, while we confirm the robustness of the Starobinsky inflation model (and other small-field models). Moreover, using unbinned CMB data, we perform a first concrete application of the GA by searching for oscillatory features in the potential in an agnostic way, and find very significant improvements upon the best featureless potentials, $\Delta {\chi }^2<-20$. These encouraging preliminary results motivate the search for resonant features in the primordial power spectrum with a multimodal distribution of frequencies. We stress that our pipeline is modular and can easily be extended to other CMB data sets and inflationary scenarios, like multifield inflation or theories with higher-order derivatives.

15 pages, 11 figures, 2 tables. Changes match published version

References in corpus (13)

• BICEP / Keck XIII: Improved Constraints on Primordial Gravitational Waves using Planck, WMAP, and BICEP/Keck Observations through the 2018 Observing Season
• A new perspective on Dark Energy modeling via Genetic Algorithms
• Primordial power spectrum from Planck
• Feature engineering and symbolic regression methods for detecting hidden physics from sparse sensors
• Uncovering the History of Cosmic Inflation from Anomalies in Cosmic Microwave Background Spectra
• Primordial Standard Clock Models and CMB Residual Anomalies
• Machine learning forecasts of the cosmic distance duality relation with strongly lensed gravitational wave events
• Regression methods in waveform modeling: a comparative study
• Euclid: constraining dark energy coupled to electromagnetism using astrophysical and laboratory data
• Euclid: Forecast constraints on consistency tests of the $\Lambda$CDM model
• Bayesian reconstruction of the inflaton's speed of sound using CMB data
• Back to the features: assessing the discriminating power of future CMB missions on inflationary models
• Cosmic Inflation and Genetic Algorithms

Cited by in corpus (9)

• Cosmological parameter estimation with Genetic Algorithms
• syren-halofit: A fast, interpretable, high-precision formula for the $\Lambda$CDM nonlinear matter power spectrum
• A precise symbolic emulator of the linear matter power spectrum
• Optimal Inflationary Potentials
• Searching for local features in primordial power spectrum using genetic algorithms
• Neural Networks for cosmological model selection and feature importance using Cosmic Microwave Background data
• Forecast constraints on null tests of the $\Lambda$CDM model with SPHEREx
• evortran: a modern Fortran package for genetic algorithms with applications from LHC data fitting to LISA signal reconstruction
• A Quantum Genetic Algorithm with application to Cosmological Parameters Estimation