Archives ouvertes de Paris-Saclay
Not a member yet
272730 research outputs found
Sort by
Signal processing and machine learning algorithms for precise timing with PICOSEC Micromegas detectors
No full textInternational audienceHigh particle rates in current and future experiments make pile-up phenomena a critical issue for extracting useful information. In this context, timing serves as a crucial fourth dimension for triggering and event reconstruction. The PICOSEC Micromegas detector has demonstrated precise timing capabilities on the order of tens of picoseconds. In this work, we develop and evaluate novel signal-processing algorithms to demonstrate the detector's potential for online precise timing. In particular, we propose an algorithm based on Artificial Neural Networks (ANNs), trained using a physically motivated model. The performance of the different algorithms is assessed using experimental data recorded during laser beam tests at the IRAMIS Facility of CEA Saclay. A timing resolution of 18.3 ± 0.6 ps, corresponding to 7.8 photoelectrons in the gas volume, is achieved — comparable to results obtained with the standard Constant Fraction Discrimination (CFD) technique. Additionally, we present an alternative algorithm that uses the integrated charge of the pulse exceeding a defined threshold as a parameter to correct for systematic effects
Euclid: Galaxy SED reconstruction in the PHZ processing function: impact on the PSF and the role of medium-band filters
No full textInternational audienceWeak lensing surveys require accurate correction for the point spread function (PSF) when measuring galaxy shapes. For a diffraction-limited PSF, as arises in space-based missions, this correction depends on each galaxy SED. In the Euclid mission, galaxy SED reconstruction, a tasks of the photometric-redshift processing function (PHZ PF), relies on broad- and medium-band ancillary photometry. The limited wavelength sampling of the Euclid VIS passband and signal-to-noise ratio may affect the reconstruction accuracy and translate into biases in the weak lensing measurements. In this study, we present the methodology, which is employed in the Euclid PHZ PF, for reconstructing galaxy SEDs at 55 wavelengths, sampling the VIS passband every 10 nm, and we assess whether it fulfils the accuracy requirements imposed on the Euclid PSF model. We employ both physics- and data-driven methods, focusing on a new approach of template-based flux correction and Gaussian processes, and we introduce an SED metric whose bias propagates into PSF quadrupole moment errors. Our findings demonstrate that Gaussian processes and template fitting meet the requirements only in specific, but complementary, redshift intervals. We therefore propose a hybrid approach, which leverages both methods. This solution proves to be effective in meeting the Euclid accuracy requirements for most of the redshift range of the survey. Finally, we investigate the impact on the SED reconstruction of a new set of 16 evenly-spaced medium-band filters for the Subaru telescope, providing quasi-spectroscopic coverage of the VIS passband. This study shows promising results, ensuring accurate SED reconstruction and meeting the mission PSF requirements. This work thus provides not only the methodological foundation of galaxy SED reconstruction in the Euclid PHZ PF, but also a roadmap for future improvements using a new medium-band survey
Euclid preparation. Testing analytic models of galaxy intrinsic alignments in the Euclid Flagship simulation
No full textInternational audienceWe model intrinsic alignments (IA) in Euclid's Flagship simulation to investigate its impact on Euclid's weak lensing signal. Our IA implementation in the Flagship simulation takes into account photometric properties of galaxies as well as their dark matter host halos. We compare simulations against theory predictions, determining the parameters of two of the most widely used IA models: the Non Linear Alignment (NLA) and the Tidal Alignment and Tidal Torquing (TATT) models. We measure the amplitude of the simulated IA signal as a function of galaxy magnitude and colour in the redshift range . We find that both NLA and TATT can accurately describe the IA signal in the simulation down to scales of -Mpc. We measure alignment amplitudes for red galaxies comparable to those of the observations, with samples not used in the calibration procedure. For blue galaxies, our constraints are consistent with zero alignments in our first redshift bin , but we detect a non-negligible signal at higher redshift, which is, however, consistent with the upper limits set by observational constraints. Additionally, several hydrodynamical simulations predict alignment for spiral galaxies, in agreement with our findings. Finally, the evolution of alignment with redshift is realistic and comparable to that determined in the observations. However, we find that the commonly adopted redshift power-law for IA fails to reproduce the simulation alignments above . A significantly better agreement is obtained when a luminosity dependence is included, capturing the intrinsic luminosity evolution with redshift in magnitude-limited surveys. We conclude that the Flagship IA simulation is a useful tool for translating current IA constraints into predictions for IA contamination of Euclid-like samples
Model calibration using Identifiability Classes: improved accuracy, robustness and speed
No full textNational audienceFinite Element Model Updating (FEMU) methods enable for the calibration of model parameters by minimizing the discrepancy between computed and measured data. However, the accuracy and reliability of these methods depend on the identifiability of the parameters, which may be challenging to determine. The recently introduced "Identifiability Classes" offer a novel framework for quantitatively evaluating parameter identifiability by combining multiple data sources, sensitivity analyses, and measurement uncertainties. Each parameter is associated with an Identifiability Class, providing insight into its sensitivity and identifiability relative to uncertainty levels, and enabling for a clear distinction between identifiable and non-identifiable parameters. This information is utilized to inform various aspects of the model updating strategy, including data source selection, temporal data selection, and regularization techniques.By analyzing the Identifiability Classes, the most relevant data source or combination of data sources is determined, with the goal of maximizing the number of identifiable parameters. Additionally, the Identifiability Classes are used to optimize the temporal resolution of simulations and measurements, by selecting the most relevant time steps with respect to the Identifiability Classes, reducing the computational costs while maintaining accuracy. When dealing with non-identifiable parameters, regularization techniques such as Tikhonov regularization are necessary. The Identifiability Classes are then used to determine the optimal level of regularization required to prevent overfitting and avoid learning noise, while still achieving accurate parameter estimates. This strategy enables for rapid updating of models while maintaining their robustness and reliability. A subsequent relaxation strategy is then employed to further refine the updated model.The application of Identifiability Classes will be demonstrated through a case study, where the constitutive parameters are updated. By selecting the most informative data sources, time steps, and regularization levels based on Identifiability Classes, accuracy and reliability of parameter estimate is improved, while reducing the computational costs.While this study focuses on constitutive parameters updating, the Identifiability Classes approach can be applied to update other FE parameters, such as boundary conditions. Furthermore, the Identifiability Classes can be used in conjunction with virtual testing to optimize experimental design, including sample geometry and loading paths, potentially reducing the number of physical experiments required while improving the speed, accuracy, and reliability of model calibration
Search for dark matter produced in association with a Higgs boson decaying to bottom quarks in proton-proton collisions at = 13 TeV
No full textInternational audienceA search for dark matter particles produced in association with a Higgs boson decaying to a bottom quark-antiquark pair in proton-proton collisions at = 13 TeV is presented. The data, collected with the CMS detector at the LHC, correspond to an integrated luminosity of 101 fb. The analysis is performed in exclusive categories targeting both Lorentz-boosted (merged) and resolved b jet pair topologies, covering a wide range of Higgs boson transverse momentum. A statistical combination is made with a previous search using data collected in 2016 and corresponding to an integrated luminosity of 35.9 fb. The observed data agree with the standard model background predictions. Constraints are placed on models predicting new particles or interactions, such as those in the simplified frameworks of baryonic-Z' and 2HDM+a, where the latter is a type-II two-Higgs-doublet model featuring a heavy pseudoscalar with an additional light pseudoscalar. Upper limits at 95% confidence level are set on the production cross section for these models. For the baryonic-Z' model, Z' boson masses below 2.25 TeV are excluded for a dark matter particle candidate mass of 1 GeV. In the 2HDM+a model, heavy pseudoscalar masses between 850 and 1300 GeV are excluded for a light pseudoscalar mass of 350 GeV
Low-uncertainty measurement of the 239Pu(n,f) cross section at n_TOF in a very broad energy range from 0.02 eV up to 10 MeV
No full textInternational audienceThe 239Pu(n,f) cross section was measured with uncertainties below 4 % from 20 meV up to 10 MeV of neutron energy at CERN’s n_TOF facility using a novel fission fragment detector with a 185.59 m flight path. Measured cross section was normalized to 1059(6) eV · b in the 9–20 eV energy range, and is consistent within uncertainties with the standard thermal value and the IAEA reference 239Pu(n,f) cross section. Experimental key values include the integral ratio I3/I1=41.20(42) defined by Duran et al. 2024 and the spectrum averaged cross section in 252Cf(sf) reference neutron field of 1802(40) mb
Rapport sur les services climatiques réussis dans le monde et leurs critères de succès: Livrable commun au Projet ciblé TRACCS-PC1-DIALOG (D5.3) et au Projet Ciblé TRACCS-PC3-DEMOCLIMA (D1)
No full textLes enjeux socio-économiques liés aux effets du réchauffement climatique sont tels qu’une demande croissante d’informations climatiques adaptées pour la mise en place de stratégies d’atténuation et/ou d’adaptation est clairement exprimée par les secteurs économiques (e.g. agriculture, énergie, tourisme, infrastructures terrestres ou maritimes, etc…) et par les territoires/régions qui mesurent pleinement leursvulnérabilités. En réponse à ces demandes, de nombreux projets de recherche nationaux (e.g., la "Convention relative à l’attribution d’un appui financier au bénéfice des services climatiques", signée entre le Ministère de la Transition Écologique et Solidaire (MTES) et le CNRS en mars 2017, ou le développement du portail DRIAS), européens (e.g., le programme ERA4CS « European Research Area for Climate Services » du JPI Climate, ou encore Copernicus) ou internationaux (e.g., les outils développés par la NASA, le GIEC…) ont permis le financement de « services climatiques ». Plusieurs acteurs opérationnels nationaux et européens développent et mettent à disposition des « services climatiques » via des plateformes souvent libres d’accès et enfin, apparaissent sur le marché des bureaux d’études dont l’activité commerciale vise le développement de « services climatiques » à la carte et adaptés aux demandes des clients.L’offre pour les « services climatiques » est donc aujourd’hui multiforme par l’information fournie (données climatiques « simples », indicateurs, outils d’aide à la décision) et par son mode de développement. Face à cette diversité, les enjeux sont de documenter et comprendre le paysage des services climatiques actuels, de faire ressortir les besoins, et de se munir de moyens pour caractériser la réussite des services climatiques, pour évaluer l'existant et guider le développement de nouveaux projets. Menée conjointement par les projets ciblés DIALOG (PC1) et DEMOCLIMA (PC3) du PEPR TRACCS, cette étude vise quatre objectifs :i) Documenter le paysage actuel des services climatiques (section 2) en France et dans le monde (identifiés par la communauté TRACCS) en les présentant par usages et cibles ;ii) Identifier des besoins non satisfaits vis-à-vis des services climatiques (section 4) ;iii) Identifier un ensemble de critères de réussite d'un service climatique pour les évaluer (section 5) ;iv) Proposer des bonnes pratiques pour atteindre ces critères de réussite (section 6); ces éléments guideront les choix des futurs démonstrateurs qui seront conçus et développés au sein de DEMOCLIMA.Notre travail repose en grande partie sur le recueil de dires d'acteurs des "services climatiques", dans une approche "bottom-up"
CausalProfiler: Generating Synthetic Benchmarks for Rigorous and Transparent Evaluation of Causal Machine Learning
No full textInternational audienceCausal machine learning (Causal ML) aims to answer "what if" questions using machine learning algorithms, making it a promising tool for high-stakes decision-making. Yet, empirical evaluation practices in Causal ML remain limited. Existing benchmarks often rely on a handful of hand-crafted or semi-synthetic datasets, leading to brittle, non-generalizable conclusions. To bridge this gap, we introduce CausalProfiler, a synthetic benchmark generator for Causal ML methods. Based on a set of explicit design choices about the class of causal models, queries, and data considered, the CausalProfiler randomly samples causal models, data, queries, and ground truths constituting the synthetic causal benchmarks. In this way, Causal ML methods can be rigorously and transparently evaluated under a variety of conditions. This work offers the first random generator of synthetic causal benchmarks with coverage guarantees and transparent assumptions operating on the three levels of causal reasoning: observation, intervention, and counterfactual. We demonstrate its utility by evaluating several state-of-the-art methods under diverse conditions and assumptions, both in and out of the identification regime, illustrating the types of analyses and insights the CausalProfiler enables
"PULSE FICTION" : Développement d'aliments et de recettes à base de légumineuses peu transformées afin de répondre aux besoins des consommateurs, des agriculteurs et d'améliorer la durabilité alimentaire
No full textInternational audiencePulses offer significant nutritional and environmental benefits and are useful components of healthier, more sustainable diets and global food security. However, their consumption in France remains low and below the world average. Farmers face economic and technical challenges in diversifying crops, and current domestic production is insufficient to meet this low demand, resulting in continued reliance on imports. Production and consumption are closely linked: low consumer demands limit incentives for farmers to expand cultivation, while limited availability can also constrain uptake. One of the major barriers to consumption relates to preparation and processing. While consumers generally view raw or minimally processed pulses positively, they often avoid using them because they perceive them as time‐consuming to prepare or associate them with undesirable sensory attributes (e.g., bitterness and astringency). It also might be difficult for consumers to categorize pulses according to their level of processing. Although pulse processing makes them more convenient for users, former studies suggest that consumers are suspicious about the processing of pulses, and that they may even confuse processing with ultra‐processing. This suspicion could sometimes lead to rejection by consumers. The process should thus remain moderate. Encouraging the use of minimally processed pulses, which balance convenience and acceptance, is crucial for wider adoption. The aim of the PULSE FICTION project is to develop minimally processed pulse products with pulses as the main ingredient that align with consumer preferences, farmer constraints and sustainability goals. A key innovation will be the selection of intermediate food products (IFPs) on the basis of consumer and farmer input and chef‐developed easy‐to‐make recipes. These products will be evaluated for their nutritional and sensory qualities, satiety, consumer acceptance and environmental impact. Beyond product development, PULSE FICTION explores the cognitive and sensory factors influencing consumer acceptability and designs effective communication strategies for all stakeholders to facilitate consumer adoption of pulse‐based foods
Global mapping of thioredoxin interacting proteins in<i>Neurospora crassa</i>
No full textThioredoxins (Trx) are essential thiol-oxidoreductases that regulate redox homeostasis by reducing oxidized cysteines in a wide range of target proteins. However, the Trx system and other redox regulation mechanisms remain poorly characterized in saprotrophic filamentous fungi. Here, we identified the components of theNeurospora crassaTrx system and uncovered potential redox-regulated proteins using Trx affinity chromatography. Genome search identified three Trx and a single thioredoxin reductase that we named TRX1, TRX2 TRX3 and TRR. Notably, TRX1 carries a C-terminal disordered extension of unknown function, conserved in two ascomycete taxa (LeotiomycetesandSordariomycetes). Using recombinant cysteine-to-serine mutants of each Trx, we performed affinity chromatography and identified 1,998 proteins - approximately 19% of theN. crassaproteome. To rank the putative Trx targets, we applied a fold enrichment metric, comparing protein abundance before and after affinity chromatography. The average fold enrichment was four, with values reaching up to 117 for the most enriched protein, a DEAD/DEAH box helicase. Among the top-enriched proteins, we identified homologs of known human and plant Trx targets, like peroxiredoxins, as well as 93 transcription factors and 38 kinases. Additional potential Trx targets encompass, ubiquitination-related enzymes, Fe-S cluster assembly proteins, phospholipases, exonucleases, and chitin synthases. Moreover, components of multiprotein complexes were co-purified, reflecting both direct Trx interactions and indirect co-association. Overall, this study provides a global map of potential redox regulated proteins and Trx targets inN. crassa, laying the ground for future investigations into redox signaling in filamentous fungi