Machine Learning the Hubble Constant

Abstract

Local measurements of the Hubble constant (H0) based on Cepheids e Type Ia supernova differ by ≈ 5σ from the estimated value of H0 from Planck CMB observations under ΛCDM assumptions. In order to better understand this H0 tension, the comparison of different methods of analysis will be fundamental to interpret the data sets provided by the next generation of surveys. In this paper, we deploy machine learning algorithms to measure the H0 through a regression analysis on synthetic data of the expansion rate assuming different values of redshift and different levels of uncertainty. We compare the performance of different algorithms as Extra-Trees, Artificial Neural Network, Extreme Gradient Boosting, Support Vector Machines, and we find that the Support Vector Machine exhibits the best performance in terms of bias-variance tradeoff, showing itself a competitive cross-check to non-supervised regression methods such as Gaussian Processes.

A thorough investigation of the prospects of eLISA in addressing the Hubble tension: Fisher forecast, MCMC and Machine Learning

“…This entails the need for faster and more efficient computational tools and data handling algorithms. Besides conventional methods of simulation and data analysis, various machine learning (ML) techniques like Gaussian Processes (GP), Genetic Algorithms (GA), and various deep learning algorithms are increasingly being used in different areas of cosmology (for a small body of diverse examples from recent years see [60][61][62][63][64][65][66][67][68][69][70][71][72][73][74][75][76][77][78]). Gaussian Processes, for example, have already found considerable application in the area of non-parametric reconstructions of various cosmological parameters [79][80][81][82].…”