arXiv Science⌕ Search
Home/Archive/arXiv

arXiv · 2610.08836

Revisiting the Hubble Tension with DESI DR2 Baryon Acoustic Oscillation Observations and Machine Learning Methods

Chenfa Zheng·Shuo Cao·Wuzheng Guo·Qiumin Wang·Xinyue Jiang·Marek Biesiada·Tonghua Liu
View original source ↗Read source document ↗Export metadata ↓

Abstract

One of the key unresolved questions in modern cosmology is the Hubble tension, which arises from the inconsistency between determinations of the Hubble constant ($H_0$) based on nearby observations and those predicted by early-Universe data under the standard $Λ$CDM paradigm. Recent advances have identified the new Dark Energy Spectroscopic Instrument (DESI) survey as a promising observational tool for probing late-time cosmology. In this paper, we use the latest DESI DR2 baryon acoustic oscillation (BAO) measurements, combined with nonparametrically reconstructed Type Ia supernovae and cosmic chronometer data, to constrain the Hubble constant using three complementary machine learning methodologies: Gaussian process regression (GPR), artificial neural networks (ANNs), and long short-term memory (LSTM) networks. We perform individual and joint constraints on $H_0$ at the effective redshifts of DESI DR2 tracers, yielding $H_0 = 61.8 \pm 5.3\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ (GPR), $66.3 \pm 5.2\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ (ANN), and $68.7 \pm 6.0\,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ (LSTM). Our results exhibit excellent agreement with the recent Planck measurements within $1σ$, which highlight the potential of data-driven, model-independent methods to address the Hubble tension. Finally, we further test the sensitivity of our results to potential systematics in the DESI data, emphasizing the importance of understanding systematic uncertainties in BAO surveys.

Explore related subjects

physics.gen-ph

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Chenfa Zheng, Shuo Cao, Wuzheng Guo, Qiumin Wang, Xinyue Jiang, Marek Biesiada, Tonghua Liu. 2026-09-29. Revisiting the Hubble Tension with DESI DR2 Baryon Acoustic Oscillation Observations and Machine Learning Methods. https://doi.org/10.3847/1538-4365%2Fae7d1a

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

An event centric approach to modeling quantum systems

We define an event-centric framework for modeling quantum systems whose only modeling elements are classical bits. An event is a correlation between two binary sequences, one representing the observer and one the observed system, and two events are related by a map, itself a sequence, through the XOR operation. We postulate that observers can access only the counts of symbols in events and maps, not their arrangement. Probabilities then follow from counting the compatible microstates, with indistinguishable microstates interfering through a $\mathbb{Z}_2$ sign, so that epistemic indeterminism and quantum spin statistics emerge without postulating Hilbert space or the Born rule. Applied to three scenarios involving spin, the model reproduces the Wigner $d$-matrix and Clebsch-Gordan probabilities in the sectors studied, with no continuous free or fitted parameters and with discrepancies that vanish as $O(1/n)$ in the sequence length $n$. In a Bell test it violates the CHSH inequality and approaches the Tsirelson bound as $n\to\infty$. Because the ontic state is attached to a pair of events, including the measurement context, the framework instantiates neither the single context-independent state space assumed by the Kochen-Specker argument nor the ontological-models framework assumed by the Pusey-Barrett-Rudolph theorem.

physics.gen-ph↗
arXiv2026

Is MOND an effect of cosmic frame-dragging à la Mach?

The physical origin of Modified Newtonian Dynamics (MOND), which predicts galaxy rotation curves via a single empirical acceleration constant $a_0$ without the need for dark matter, remains unknown. I propose a modified-inertia origin for both MOND's interpolating function and $a_0$ within a framework closely related to Gravitoelectromagnetism, the linearized limit of General Relativity. Following Sciama's model unifying gravity and inertia, which implements Mach's principle and the relativity of acceleration through gravitoelectric induction, I consider an idealized homogeneous universe undergoing constant Hubble expansion. I identify $a_0 \sim H_0^2R_u \sim c^2/R_u$ with the relative recession acceleration of cosmic matter near the horizon, which dominates the cosmological contribution. I show that this scale enters local inertial dynamics only when the relative acceleration of the universe becomes anisotropic and the Rindler horizon is comparable to or larger than the cosmic Hubble horizon, at accelerations of order $a_0$ and below. By the simplest Ansatz for the observer's acceleration-dependent causal volume, accounting for the interplay between the cosmic and Rindler horizons, I derive MOND's simple interpolating function as a modified inertia law without the need to insert it by hand or make any further ad hoc adjustments. Physically, the region brought into causal connection ahead of the accelerating observer by the deformation of the cosmic horizon contributes less inertial drag than is lost from the region removed behind him by the Rindler horizon. This proof of concept motivates the possibility that MOND arises from large-scale linear inertial frame-dragging induced by the expanding universe within General Relativity.

physics.gen-ph↗
arXiv2026

On the Meaning of Localization in Non-Local Quantum Field Theory

In this paper we explore and derive an uncertainty principle for an ultraviolet complete nonlocal quantum field theory where under our hypothesises of an induced equal time detector response kernel, we then prove that the observed localization width obeys an exact variance addition law. Then when we combine this with the ordinary Heisenberg inequality and we obtain a nonlocal uncertainty relation. The bound reduces to the usual local relation in the infrared or local limit when $E_M \to \infty$, while in the ultraviolet it implies a minimal localization length of order $L_M$. We go on to explain what this means for locality, microcausality, the interpretation of spacetime points, and the ultraviolet structure of quantum field theory. In this formulation we note and prove that spacetime will remain a Lorentz covariant continuum at the level of the manifold description but pointlike localization ceases to be a physically realizable observable notion below the nonlocality scale.

physics.gen-ph↗