Publications
-
Effet de la complexité du réseau LSTM sur l’explicabilité en Maintenance Prédictive
La nature complexe des données en maintenance prédictive impose souvent l’utilisation de modèles d’apprentissage profonds. Malgré leur efficacité dans la prédiction du RUL (durée de vie résiduelle des machines), ces « boites noires » fournissent des résultats qui ne sont pas directement compréhensibles. Ainsi, des méthodes XAI post hoc sont gé néralement utilisées pour les […]
-
Hyperparameter Impact on Computational Efficiency in Federated Edge Learning
The heterogeneity induced by the federated edge learning execution environment poses many performance challenges. Indeed, a balance between efficient resource usage and inference accuracy must be found. Our work therefore aims at characterizing the hyperparameter influence by creating a variety of simulated execution circumstances. We designed an experimentation platform to simulate the execution of a […]
-
Optimizing Shapley Value for Client Valuation in Federated Learning through Enhanced GTG-Shapley
In the ever-evolving realm of federated learning (FL), the question of data worth resonates with newfound urgency across organizations and individuals. In the dynamic FL ecosystem, where data resides across distributed nodes, evaluating the value of each client’s data is paramount. The evaluation mechanism helps to understand individual contributions to the overall process and incentivizes […]
-
Enhancing Explainability in Predictive Maintenance: Investigating the Impact of Data Preprocessing Techniques on XAI Effectiveness
In predictive maintenance, the complexity of the data often requires the use of Deep Learning models. These models, called ”black boxes”, have proved their worth in predicting the Remaining Useful Life (RUL) of industrial machines . However, the inherent opacity of these models requires the incorporation of post-hoc explanation methods to enhance transparency. The quality […]
-
GDLC: A new Graph Deep Learning framework based on centrality measures for intrusion detection in IoT networks
The increasing growth of the Internet of Things (IoT) with the diverse and dynamic nature of devices made detecting and preventing network intrusions more important and challenging. As new and sophisticated cyber-attacks are being used, there is an increasing need for advanced intrusion detection systems that can adapt to emerging threats. The majority of existing […]
-
An intelligent mechanism for energy consumption scheduling in smart buildings
In recent years, the incorporation of sensing technology into residential buildings has given rise to the concept of ”smart buildings”, aimed at enhancing resident comfort. These buildings are typically part of interconnected neighborhoods sharing common energy sources, which makes the energy consumption a critical consideration in decision-making processes. Consequently, optimizing energy usage in smart buildings […]
-
Multi-Task Learning for PBFT Optimisation in permissioned Blockchains
Finance, supply chain, healthcare, and energy have an increasing demand for secure transactions and data exchange. Permissioned blockchains fulfilled this need thanks to the consensus protocol that ensures that participants agree on a common value. One of the most widely used protocols in private blockchains is the Practical Byzantine Fault Tolerance (PBFT) which tolerates up […]
-
Analyse de sensibilité du modèle énergétique d’un bâtiment pédagogique constitué de conteneurs maritimes
Les modèles énergétiques, bien que très utiles pour prédire le comportement physique du bâtiment, comportent des incertitudes (paramètres inconnus, approximations…). Ces incertitudes ont un impact sur la fiabilité des prédictions, qu’il est possible de limiter via la calibration du modèle. Nous présentons dans cet article une modélisation énergétique d’un bâtiment pédagogique construit à partir de […]
-
Apprentissage automatique d’EDPs contraintes par la physique pour l’identification des hétérogénéités dans les structures mécaniques élancées
On propose une méthode d’identification automatique d’EDPs hétérogènes en espace dont l’application visée est le contrôle non-destructif des structures mécaniques à partir de sollicitations en dynamique basse fréquence. Cette méthode reprend le principe d’approches récentes de construction automatique de modèles mathématiques en introduisant la possibilité d’intégrer des hétérogénéités en espace. Le problème d’identification résultant est […]
-
Maximizing the number of satisfied charging demands of electric vehicles on identical chargers
This paper addresses the electric vehicle charging scheduling problem in a charging station with a limited overall power capacity and a limited number of chargers. Electric vehicle drivers submit their charging demands. Given the limited resources, these charging demands are either accepted or rejected and accepted demands must be satisfied. The objective of the scheduler […]
-
Simultaneous scheduling of production and maintenance in a flexible job-shop workshop
This paper proposes an integer linear program for joint and simultaneous scheduling of production and predictive maintenance tasks based on degradation prediction. The objective is to find a scheduling scheme that minimizes the total execution time while guaranteeing that the state of the machines is maintained below a predefined degradation threshold. Experiments are performed on […]
-
Exploring the Capacitated Vehicle Routing Problem using the power of Machine Learning : A Literature Review
The Capacitated Vehicle Routing Problem is a classic logistics optimization challenge for which numerous approaches are employed to address it. A prevalent strategy involves employing machine learning for predicting demand, leveraging historical data to accurately forecast customer needs. Furthermore, reinforcement learning techniques are employed to optimize vehicle routes, adapting them dynamically in response to changing […]
Loading…
Erreur : tout le contenu a été chargé.