When: Thursday 20th March 3:30 pm 

Where: Telecom SudPAris Evry Room E203  

Title: Data Pipeline System for In-Network Distributed Learning

Abstract: Beyond 5G technology aims to fully automate network reconfiguration using AI-based systems to consistently ensure user service quality . The network analysis is essential to achieve this goal. Relying solely on past network data is memory-intensive, as it requires storing vast amounts of data collected from distributed network components and may fail to fully capture the current state of the network. In this talk, we introduce a data pipeline system that leverages distributed learning to support (near) real-time network analysis for anomaly detection.

Short Bio: Dr. Patient NTUMBA is a postdoctoral researcher in the Networks and Connected Objects (ROC) team at the CEDRIC laboratory, CNAM, Paris. He completed his Ph.D. with the Middleware on the Move (MiMove) team at the INRIA Paris Center. His research focuses on challenges related to data collection, transfer, and processing in the emerging networked distributed systems such as IoT, Edge/Fog computing, Cloud computing, and 5G. The primary objective of his work is to ensure user service quality while optimizing the use of network, computational, and energy resources across these networked distributed systems.

Contact pour plus d’information:Vincent Gauthier