Constraints on a dark energy dominated Universe are obtained from an interplay between Bayesian Machine Learning and string Swampland criteria. The approach here differs from previous studies, since in the generative process Swampland criteria are used and, only later, the results of the fit are validated, by using observational data-sets. A generative process based Bayesian Learning approach is applied to two models and the results are validated by means of available $H(z)$ data. For the first model, a parametrization of the Hubble constant is considered and, for the second, a parametrization of the deceleration parameter. This study is motivated by a recent work, where constraints on string Swampland criteria have been obtained from a Gaussian Process and $H(z)$ data. However, the results obtained here are fully independent of the observational data and allow to estimate how the high-redshift behavior of the Universe will affect the low-redshift one. Moreover, both parameterizations in the generative process, for the Hubble and for the deceleration parameters, are independent of the dark energy model. The outcome, both data- and dark energy model-independent, may highlight, in the future, the borders of the Swampland for the low-redshift Universe and help to develop new string-theory motivated dark-energy models. The string Swampland criteria considered might be in tension with recent observations indicating that phantom dark energy cannot be in the Swampland. Finally, a spontaneous sign switch in the dark energy equation of state parameter is observed when the field traverses are in the $z\in[0,5]$ redshift range, a remarkable phenomenon requiring further analysis.

Bayesian Learning and elements of probabilistic programming are used to probe the opacity of the Universe, as a way of addressing the $H_{0}$ tension issue. Bayesian Learning allows to test the models well beyond the limits imposed by actual astronomical observations. In this approach, the model is the key object to generate the data, and the generative process involves the unknown parameters of the model, what permits to constrain them, too. In this way, prior beliefs of the unknown parameters can be incorporated, and be used to get the posterior results, and constrain them appropriately, in this way. Using Bayesian learning algorithms, beliefs on the parameters are thus updated, and a new distribution over them results. In our study, constraints on the cosmic opacity are determined for two flat models, $\Lambda$CDM and XCDM (this having $\omega_{de} \neq -1$), for three redshift ranges, $z\in[0,2.5]$, $z \in[0,5]$, and $z\in[0,10]$, in each case. The following forms for the opacity, $\tau(z) = 2\epsilon z$ and $\tau(z) = (1+z)^{2\epsilon} -1$, corresponding to an observer at $z=0$ and a source at $z$, are considered. The results of our analysis show that, very probably, the Universe is not fully transparent, and this may have a significant impact on the dark energy equation of state parameter and on the $H_{0}$ tension problem. In the generative process, the fact that the flux received by the observer is reduced, owing to cosmic opacity, has been taken into account, and the observed luminosity distance turns out to be: $D_{L,obs} = D_{L,true} e^{\tau/2}$, being $\tau$ the opacity parameter, and $D_{L,true}$ the luminosity distance associated to the cosmological model employed.

In this paper, two inhomogeneous single fluid models for the Universe, which are able to naturally solve the $H_{0}$ tension problem, are discussed. The analysis is based on a Bayesian Machine Learning approach that uses a generative process. The adopted method allows to constrain the free parameters of each model by using the model itself, only. The observable is taken to be the Hubble parameter, obtained from the generative process. Using the full advantages of our method, the models are constrained for two redshift ranges. Namely, first this is done with mock $H(z)$ data over $z\in [0,2.5]$, thus covering known $H(z)$ observational data, which are most helpful to validate the fit results. Then, aiming to extend to redshift ranges to be covered by the most recent ongoing and future planned missions, the models are constrained for the range $z\in[0,5]$, too. Full validation of the results for this extended redshift range will have to wait for the near future, when higher redshift $H(z)$ data become available. This makes our models fully falsifiable. Finally, our second model here is able to explain the BOSS reported value for $H(z)$ at $z=2.34$.

Energy flows between dark energy and dark matter may alleviate the Hubble tension and mitigate the coincidence problem. In this paper, after reconstructing the redshift evolution of the Hubble function by adopting Gaussian process techniques, we estimate the best-fit parameters for some flat Friedmann cosmological models based on a Modified Chaplygin Gas interacting with dark matter. In fact, the expansion history of the Universe will be investigated because passively evolving early galaxies constitute cosmic chronometers. An estimate for the present-day values of the deceleration parameter, adiabatic speed of sound within the dark energy fluid, effective dark energy, and dark matter equation of state parameters is provided. By this, we mean that the interaction term between the two dark fluids, which breaks the Bianchi symmetries, will be interpreted as an effective contribution to the dark matter pressure similarly to the framework of the \lq\lq Generalized Dark Matter". Fixing a certain value for the dark matter abundance at the present day as initial condition will allow us to investigate whether the estimate of the Hubble constant is sensitive to the dark matter - dark energy coupling. We will also show that the cosmic chronometers data favor a hot dark matter, and that our findings are in agreement with the Le Ch\^atelier-Braun principle according to which dark energy should decay into dark matter (and not vice versa).

We apply Gaussian processes and Hubble function data in $f(T)$ cosmology, to reconstruct for the first time the $f(T)$ form in a model-independent way. In particular, using $H(z)$ datasets coming from cosmic chronometers as well as from the radial BAO method, alongside the latest released local value $H_{0} = 73.52 \pm 1.62$ km/s/Mpc, we reconstruct $H(z)$ and its derivatives, resulting eventually in a reconstructed region for $f(T)$, without any assumption. Although the cosmological constant lies in the central part of the reconstructed region, the obtained mean curve follows a quadratic function. Inspired by this we propose a new $f(T)$ parametrization, i.e. $f(T) = -2\Lambda +\xi T^2$, with $\xi$ the sole free parameter that quantifies the deviation from $\Lambda$CDM cosmology. Additionally, we confront three viable one-parameter $f(T)$ models of the literature, which respectively are the power-law, the square-root exponential, and the exponential one, with the reconstructed $f(T)$ region, and then we extract significantly improved constraints for their model parameters, comparing to the constraints that arise from usual observational analysis. Finally, we argue that since we are using the direct Hubble measurements and the local value for $H_0$ in our analysis, with the above reconstruction of $f(T)$, the $H_0$ tension can be efficiently alleviated.

In this paper we study implications of the possible excess of 21-cm line global signal at the epoch of cosmic dawn on the evolutions of a class of dynamically interacting dark energy (IDE) models. We firstly summarize two dynamical mechanisms in which different background evolutions can exert considerable effects on the 21-cm line global signal. One is the decoupling time of Compton scattering heating, the other stems from the direct change of optical depth due to the different expansion rate of the Universe. After that, we investigate the IDE models to illustrate the tension between the results of Experiment to Detect the Global Epoch of reionization Signature (EDGES) and other experiments. To apply the analyses of these two mechanisms to IDE models, we find that only the optical depth can be significantly changed. Accordingly, in order to relieve the tension by including the effects of the decoupling time of Compton scattering heating, we deduce a possible evolution form for the Hubble parameter within IDE that begins at an early stage around $z\sim 100$ and then smoothly evolves to a value at $z\sim 17$ which is smaller than that obtained in the standard paradigm. Eventually, we fulfill this scenario by adding an early dark energy dominated stage to the cosmological paradigm described by IDE models, which can alleviate the tension between EDGES and other cosmological observations but cannot completely solve it.

Implications of the string Swampland criteria for a dark energy dominated universe, obtained by using Gaussian processes and $H(z)$ data analysis, are discussed. In particular, the Swampland criteria for a scalar-field dark energy, without assuming any specific form for the potential. By allowing the Gaussian process to reconstruct the form of the potential from $H(z)$ data, upper bounds on the second Swampland criterion (involving $|V^{\prime}|/V$) for two different kernel functions (the squared exponential and Matern~($\nu = 9/2$) kernels) are estimated . The approach here differs from previous studies, since the upper bound of the second Swampland criterion is derived in a thoroughly model-independent way, without resorting to a model-to-model comparison strategy. The analysis is performed using the latest values of $H_{0}$ reported by the Planck and Hubble missions. Results for the estimation of the constant of $SC2$ hint towards the possibility of getting upper bounds well behind the estimations for the dark energy dominated universe as reported in previous studies, which involved model-to-model comparison. Estimations from this new approach turn to be quite sensitive and just depend on the quality of the data and on the kernel employed. This study is a first attempt towards the exploitation of the Swampland criteria in a model-independent way and may be extended to involve other datasets and, also, in trying to understand what is the impact of higher-redshift data on the upper bounds. In the analysis, $40$-point $H(z)$ data have been used, consisting of a $30$-point sample deduced from a differential age method and an additional $10$-point sample obtained from the radial BAO method. Hints towards the possibility of eventually disproving the Swampland conjecture are noted.

Future singularities arising in a family of models for the expanding Universe, characterized by sharing a convenient parametrization of the energy budget in terms of the deceleration parameter, are classified. Finite-time future singularities are known to appear in many cosmological scenarios, in particular, in the presence of viscosity or non-gravitational interactions, the last being known to be able to suppress or just change in some cases the type of the cosmological singularity. Here, a family of models with a parametrization of the energy budget in terms of the deceleration parameter are studied in the light of Gaussian processes using reconstructed data from $40$-value $H(z)$ datasets. Eventually, the form of the possible non-gravitational interaction between dark energy and dark matter is constructed from these smoothed $H(z)$ data. Using phase space analysis, it is shown that a non-interacting model with dark energy $\omega_\mathrm{de} = \omega_{0} + \omega_{1}q$ ($q$ being the deceleration parameter) may evolve, after starting from a matter dominated unstable state, into a de Sitter Universe (the solution being in fact a stable node). Moreover, for a model with interaction term $Q = 3 H b \rho_\mathrm{dm}$ ($b$ is a parameter and $H$ the Hubble constant) three stable critical points are obtained, what may have important astrophysical implications. In addition, part of the paper is devoted to a general discussion of the finite-time future singularities obtained from direct numerical integration of the field equations, since they appear in many cosmological scenarios and could be useful for future extended studies of the models here introduced. Numerical solutions for the new models, produce finite-time future singularities of Type I or Type III, or an $\omega$-singularity, provided general relativity describes the background dynamics.