Related papers: BBNet: accurate neural network emulator for primor…
Deep Neural Networks (DNNs) are powerful tools for various computer vision tasks, yet they often struggle with reliable uncertainty quantification - a critical requirement for real-world applications. Bayesian Neural Networks (BNN) are…
This paper is superseded by Arxiv:1911.07334. Big-bang nucleosynthesis (BBN) is a valuable tool to constrain the physics of the early universe and is the only probe of the radiation-dominated epoch. A fundamental assumption in BBN is that…
Big bang nucleosynthesis (BBN) and the cosmic microwave background (CMB) anisotropies independently predict the universal baryon density. Comparing their predictions will provide a fundamental test on cosmology. Using BBN and the CMB…
The relic abundances of the light elements synthesized during the first few minutes of the evolution of the Universe provide unique probes of cosmology and the building blocks for stellar and galactic chemical evolution, while also enabling…
Big bang nucleosynthesis (BBN), an epoch of primordial nuclear transformations in the expanding Universe, has left an observable imprint in the abundances of light elements. Precision observations of such abundances, combined with…
We study effects of long-lived massive particles, which decay during the big-bang nucleosynthesis (BBN) epoch, on the primordial abundances of light elements. Compared to the previous studies, (i) the reaction rates of the standard BBN…
Limits can be placed on nonstandard neutrino physics when big bang nucleosynthesis (BBN) calculations employing standard neutrino physics agree with the observationally inferred primordial abundances of deuterium (D), $^3$He, $^4$He, and…
Standard big bang nucleosynthesis (BBNS) promises accurate predictions of the primordial abundances of deuterium, helium-3, helium-4 and lithium-7 as a function of a single parameter. Previous measurements have nearly always been…
We present the first public release of our generic neural network training algorithm, called SkyNet. This efficient and robust machine learning tool is able to train large and deep feed-forward neural networks, including autoencoders, for…
The primordial abundances of the light nuclides produced by Big Bang Nucleosynthesis (BBN) during the first 20 minutes in the evolution of the Universe are sensitive to the universal density of baryons and to the expansion rate of the early…
Deep neural networks (DNNs) have achieved great success in the area of computer vision. The disparity estimation problem tends to be addressed by DNNs which achieve much better prediction accuracy in stereo matching than traditional…
We investigate the use of approximate Bayesian neural networks (BNNs) in modeling hundreds of time-delay gravitational lenses for Hubble constant ($H_0$) determination. Our BNN was trained on synthetic HST-quality images of strongly lensed…
Binarized neural networks, or BNNs, show great promise in edge-side applications with resource limited hardware, but raise the concerns of reduced accuracy. Motivated by the complex neural networks, in this paper we introduce complex…
Deep learning with artificial neural networks is increasingly gaining attention, because of its potential for data-driven astronomy. However, this methodology usually does not provide uncertainties and does not deal with incompleteness and…
We study the impact of an early dark energy component (EDE) present during big bang nucleosynthesis (BBN) on the elemental abundances of deuterium D/H, and helium $Y_p$, as well as the effective relativistic degrees of freedom $N_{\rm…
Using a new sample of extremely metal poor systems, the EMPRESS survey has recently reported a primordial helium abundance that is $3\sigma$ smaller than the prediction from the standard big bang nucleosynthesis (BBN) scenario. This…
Bayesian networks (BNs) are attractive, because they are graphical and interpretable machine learning models. However, exact inference on BNs is time-consuming, especially for complex problems. To improve the efficiency, we propose a fast…
Motivated by the recent measurement of the primeval abundance of deuterium, we re-examine the nuclear inputs to big-bang nucleosynthesis (BBN). Using Monte-Carlo realization of the nuclear cross-section data to directly estimate the…
We assess the effects on Big Bang Nucleosynthesis (BBN) of lepton number generation in the early universe resulting from the two-doublet four-neutrino mass/mixing scheme. It has been argued that this neutrino mass/mixing arrangement gives…
Of the light nuclides observed in the universe today, D, 3He, 4He, and 7Li are relics from its early evolution. The primordial abundances of these relics, produced via Big Bang Nucleosynthesis (BBN) during the first half hour of the…