Fully Asynchronous Distributed Learning with Adversaries: Theory and Applications
Fully Asynchronous Distributed Learning with Adversaries: Theory and Applications
Fully Asynchronous Distributed Learning with Adversaries: Theory and Applications is a collaborative project composed of two research teams:
Indian Institute of Sciences (India)
IMT Atlantique (France)
The project officially started in July 2024, being financially supported for 3 years up until June 2027.
Goal: Design distributed stochastic optimization algorithms that are:
asynchronous
robust to network faults and malicious attacks
fast
communication efficient
differentially private.
IMT Atlantique
Alexandre Reiffers-Masson (Associate Professor, IMT Atlantique)
Vishal Halder (Ph.D., IMT Atlantique)
Indian Institute of Science
Gugan Thoppe (Assistant Professor, IISc)
T. Nagesh (Postdoctoral fellow, IISc)
Nibedita Roy (Ph.D., IISc)
Visits from India to France:
Gugan Thoppe visits IMT Atlantique from 03/07/2024 - 21/07/2024
Nibedita Roy - T. Nagesh visits IMT Atlantique from 26/05/2025-24/06/2025
Visits from France to India:
Alexandre Reiffers-Masson visits IISc from 05/05/2025-09/05/2025
Papers
In 2024, we have published the following papers:
REF + ABSTRACT