A5087628438- Radiative Heat Transfer Studies
- Neural dynamics and brain function
- Calibration and Measurement Techniques
- Advanced Memory and Neural Computing
- Neural Networks and Applications
- Neural Networks and Reservoir Computing
- Atmospheric chemistry and aerosols
- Atmospheric aerosols and clouds
- Atmospheric and Environmental Gas Dynamics
- Combustion and flame dynamics
- Wind and Air Flow Studies
- EEG and Brain-Computer Interfaces
- Atmospheric Ozone and Climate
- Privacy-Preserving Technologies in Data
- Computational Drug Discovery Methods
- Visual perception and processing mechanisms
- Spectroscopy and Laser Applications
- advanced mathematical theories
- Neuroscience and Neural Engineering
- Stochastic processes and financial applications
- Sleep and Wakefulness Research
- Optical Imaging and Spectroscopy Techniques
- Breast Cancer Treatment Studies
- Biosimilars and Bioanalytical Methods
- Non-Invasive Vital Sign Monitoring
Centre National de la Recherche Scientifique
2013-2024
Centre d'Énergétique et de Thermique de Lyon
2013-2024
Université Claude Bernard Lyon 1
2013-2024
Institut National des Sciences Appliquées de Lyon
2020-2024
IMT Mines Albi
2013-2023
Daikin (United States)
2021
Institut des Neurosciences Paris-Saclay
2018
Unit of Neuroscience Information and Complexity
2016-2018
Université Paris-Sud
2018
Laboratoire Dynamiques Sociales et Recomposition des Espaces
2018
Sleep stage classification constitutes an important preliminary exam in the diagnosis of sleep disorders. It is traditionally performed by a expert who assigns to each 30 s signal stage, based on visual inspection signals such as electroencephalograms (EEGs), electrooculograms (EOGs), electrocardiograms, and electromyograms (EMGs). We introduce here first deep learning approach for that learns end-to-end without computing spectrograms or extracting handcrafted features, exploits all...
Federated multipartner machine learning has been touted as an appealing and efficient method to increase the effective training data volume thereby predictivity of models, particularly when generation is resource-intensive. In landmark MELLODDY project, indeed, each ten pharmaceutical companies realized aggregated improvements on its own classification or regression models through federated learning. To this end, they leveraged a novel implementation extending multitask across partners,...
Recent research has shown that auditory closed-loop stimulation can enhance sleep slow oscillations (SO) to improve N3 quality and cognition. Previous studies have been conducted in lab environments. The present study aimed validate assess the performance of a novel ambulatory wireless dry-EEG device (WDD), for SO during at home. WDD detect automatically send on were tested 20 young healthy subjects who slept with both miniaturized polysomnography (part 1) stimulated sham nights within...
To apply federated learning to drug discovery we developed a novel platform in the context of European Innovative Medicines Initiative (IMI) project MELLODDY (grant n°831472), which was comprised 10 pharmaceutical companies, academic research labs, large industrial companies and startups. The first industry-scale enable creation global model for without sharing confidential data sets individual partners. trained on by aggregating gradients all contributing partners cryptographic, secure way...
It was recently shown that radiation, conduction and convection can be combined within a single Monte Carlo algorithm such an immediately benefits from state-of-the-art computer-graphics advances when dealing with complex geometries. The theoretical foundations make this coupling possible are fully exposed for the first time, supporting intuitive pictures of continuous thermal paths run through different physics at work. First, frameworks propagators Green's functions used to demonstrate...
Urban overheating, driven by the increasing frequency of extreme heat events, poses significant challenges to public health and thermal comfort in densely populated areas. Factors such as inadequate building design, a lack accessible outdoor shelters further intensify these challenges, underscoring urgent need for urban adaptation strategies.Green infrastructure, particularly continuous tree cover parks, effectively mitigates stress environments through shading evapotranspiration processes....
Machine learning is promising, but it often needs to process vast amounts of sensitive data which raises concerns about privacy. In this white-paper, we introduce Substra, a distributed framework for privacy-preserving, traceable and collaborative Learning. Substra gathers providers algorithm designers into network nodes that can train models on demand under advanced permission regimes. To guarantee privacy, implements learning: the never leave their nodes; only algorithms, predictive...
Federated multi-partner machine learning can be an appealing and efficient method to increase the effective training data volume thereby predictivity of models, particularly when generation is resource intensive. In landmark MELLODDY project, each ten pharmaceutical companies realized aggregated improvements on its own classification and/or regression models through federated learning. To this end, they leveraged a novel implementation extending multi-task across partners, platform audited...
Monte Carlo is famous for accepting model extensions and refinements up to infinite dimension. However, this powerful incremental design based on a premise which has severely limited its application so far: state-variable can only be recursively defined as function of underlying state-variables if linear. Here we show that alleviated by projecting nonlinearities onto polynomial basis increasing the configuration space Considering phytoplankton growth in light-limited environments, radiative...
Identifying, formalizing, and combining biological mechanisms that implement known brain functions, such as prediction, is a main aspect of research in theoretical neuroscience. In this letter, the spike-timing-dependent plasticity homeostatic plasticity, combined an original mathematical formalism, are shown to shape recurrent neural networks into predictors. Following rigorous treatment, we prove they online gradient descent distance between network activity its stimuli. The convergence...
This paper deals with the application of temporal averaging methods to recurrent networks noisy neurons undergoing a slow and unsupervised modification their connectivity matrix called learning. Three time-scales arise for these models: (i) fast neuronal dynamics, (ii) intermediate external input system, (iii) learning mechanisms. Based on this time-scale separation, we apply an extension mathematical theory stochastic periodic forcing in order derive reduced deterministic model dynamics. We...