Publicaciones

Affichage de 181 à 190 sur 5539


  • Communication dans un congrès

Accelerated NAS via pretrained ensembles and multi-fidelity Bayesian Optimization

Houssem Ouertatani, Cristian Maxim, Smail Niar, El-Ghazali Talbi

Bayesian optimization (BO) is a black-box search method particularly valued for its sample efficiency. It is especially effective when evaluations are very costly, such as in hyperparameter optimization or Neural Architecture Search (NAS). In this work, we design a fast NAS method based on BO....

33rd International Conference on Artificial Neural Networks (ICANN), Sep 2024, Lugano, Switzerland. ⟨10.1007/978-3-031-72332-2_17⟩. ⟨hal-04611343⟩

  • Article dans une revue

Underbody flow control for base drag reduction of a real car model

Laurent Keirsbulck, Olivier Cadot, Marc Lippert, David Boussemart, Jérémy Basley, Sébastien Delprat, Sébastien Paganelli

A 1:5 scale realistic car model of the original Twingo GT but presenting a flat underbody and no exhaust line is tested in a wind tunnel at Reynolds numbers Re = 2.15 × 10^5 to 4.3 × 10^5. A range of underbody flow characteristic velocities Ub = [0.5 − 0.72]U∞ (U∞ the free-stream velocity) is...

Journal of Wind Engineering and Industrial Aerodynamics, 2024, 252, pp.105822. ⟨10.1016/j.jweia.2024.105822⟩. ⟨hal-04819727⟩

  • Article dans une revue

Deep reinforcement learning with predictive auxiliary task for autonomous train collision avoidance

Antoine Plissonneau, Luca Jourdan, Damien Trentesaux, Lotfi Abdi, Mohamed Sallak, Abdelghani Bekrar, Benjamin Quost, Walter Schön

The contribution of this paper consists of a deep reinforcement learning (DRL) based method for autonomous train collision avoidance. While DRL applied to autonomous vehicles’ collision avoidance has shown interesting results compared to traditional methods, train-like vehicles are not currently...

Journal of Rail Transport Planning & Management, 2024, 31, pp.100453. ⟨10.1016/j.jrtpm.2024.100453⟩. ⟨hal-04607863⟩