An approach to cold dark matter deviation and the $H_{0}$ tension problem by using machine learning

Abstract

In this work, two different models, one with cosmological constant Λ, and baryonic and dark matter (with ω dm = 0), and the other with an X dark energy (with ω de = −1), and baryonic and dark matter (with ω dm = 0), are investigated and compared. Using Bayesian machine learning analysis, constraints on the free parameters of both models are obtained for the three redshift ranges: z ∈ [0, 2], z ∈ [0, 2.5], and z ∈ [0, 5], respectively. For the first two redshift ranges, high-quality observations of the expansion rate H(z) exist already, and they are used for validating the fitting results. Additionally, the extended range z ∈ [0, 5] provides predictions of the model parameters, verified when reliable higher-redshift H(z) data are available. This learning procedure, based on the expansion rate data generated from the background dynamics of each model, shows that, at cosmological scales, there is a deviation from the cold dark matter paradigm, ω dm = 0, for all three redshift ranges. The results show that this approach may qualify as a solution to the H0 tension problem. Indeed, it hints at how this issue could be effectively solved (or at least alleviated) in cosmological models with interacting dark energy.

Solving the $H_{0}$ tension in $f(T)$ Gravity through Bayesian Machine Learning

“…This can be done by challenging not only our understanding of dark energy but also the dark matter part. To this point, recently, it has been demonstrated using BML that there is a deviation from the cold dark matter paradigm on cosmological scales, which might efficiently solve the H 0 tension [41].…”

“…This important tension motivated researchers to look for different solutions ranging from indications of new physics to possible hidden sources of systematic errors, and biases in observational data [38][39][40][41] (see references therein for other options to solve the H 0 tension). Indeed, to understand the source of such discrepancy that can challenge the ΛCDM model, other independent observational sources have been used to to determine the value of H 0 .…”

Reconstruction of the quintessence dark energy potential from a Gaussian process

“…The fact that ML is designed to find the questions from the answers allows to hope that, in the near future, interesting developments in this direction may arise. We would like to mention that there is another interesting approach, known as Bayesian Machine Learning, which we hope can eventually be very efficient in overcoming such limitations [4,10,11] (see there how it can be used to tackle the H 0 tension problem). Now, let us come back to Ref.…”

Viscous coupled fluids in terms of a log-corrected equation of state

“…Such an approach was recently outlined in Ref. [17], where two models were investigated: (i) cosmological constant Λ + baryonic matter + dark matter (with ω dm = 0), and (ii) dark energy (with ω de = −1) + baryonic matter + dark matter (with ω dm = 0). Comparison with experimental data gave, as a main result, that ω dm = 0 was favored in both cases, thus indicating a deviation from the usual cold dark matter model.…”

“…Our approach is thus relying upon the ansatz (1), and is more theoretical in nature than that followed in Ref. [17]. We assume a spatially flat Friedmann-Robertson-Walker space-time, and assume homogeneity and isotropy at a large scale.…”