arXiv subjects

Marion Ullmo

Publications and source records attributed to Marion Ullmo.

3 recordsLinked to original sources

arXiv2024

Nonparametric signal separation in very-high-energy gamma ray observations with probabilistic neural networks

An intriguing challenge in observational astronomy is the separation signals in areas where multiple signals intersect. A typical instance of this in very-high-energy (VHE, E$\gtrsim$100 GeV) gamma-ray astronomy is the issue of residual background in observations. This background arises when cosmic-ray protons are mistakenly identified as gamma-rays from sources of interest, thereby blending with signals from astrophysical sources of interest. We introduce a deep ensemble approach to determine a non-parametric estimation of source and background signals in VHE gamma observations, as well as a likelihood-derived epistemic uncertainty on these estimations. We rely on minimal assumptions, exploiting the separability of space and energy components in the signals, and defining a small region in coordinate space where the source signal is assumed to be negligible compared to background signal. The model is applied both on mock observations, including a simple toy case and a realistic simulation of dark matter annihilation in the Galactic center, as well as true observations from the public H.E.S.S. data release, specifically datasets of the Crab nebula and the pulsar wind nebula MSH 15-52. Our method performs well in mock cases, where the ground truth is known, and compares favorably against conventional physical analysis approaches when applied to true observations. In the case of the mock dark matter signal in the Galactic center, our work opens new avenues for component separation in this complex region of the VHE sky.

astro-ph.HE↗

arXiv2024

Predicting large scale cosmological structure evolution with generative adversarial network-based autoencoders

Predicting the nonlinear evolution of cosmic structure from initial conditions is typically approached using Lagrangian, particle-based methods. These techniques excel in terms of tracking individual trajectories, but they might not be suitable for applications where point-based information is unavailable or impractical. In this work, we explore an alternative, field-based approach using Eulerian inputs. Specifically, we developed an autoencoder architecture based on a generative adversarial network (GAN) and trained it to evolve density fields drawn from dark matter N-body simulations. We tested this method on both 2D and 3D data. We find that while predictions on 2D density maps perform well based on density alone, accurate 3D predictions require the inclusion of associated velocity fields. Our results demonstrate the potential of field-based representations to model cosmic structure evolution, offering a complementary path to Lagrangian methods in contexts where field-level data is more accessible.

astro-ph.CO↗

arXiv2020

Encoding large scale cosmological structure with Generative Adversarial Networks

Recently a type of neural networks called Generative Adversarial Networks (GANs) has been proposed as a solution for fast generation of simulation-like datasets, in an attempt to bypass heavy computations and expensive cosmological simulations to run in terms of time and computing power. In the present work, we build and train a GAN to look further into the strengths and limitations of such an approach. We then propose a novel method in which we make use of a trained GAN to construct a simple autoencoder (AE) as a first step towards building a predictive model. Both the GAN and AE are trained on images issued from two types of N-body simulations, namely 2D and 3D simulations. We find that the GAN successfully generates new images that are statistically consistent with the images it was trained on. We then show that the AE manages to efficiently extract information from simulation images, satisfyingly inferring the latent encoding of the GAN to generate an image with similar large scale structures.

astro-ph.CO↗