What every Distributed Computing Researcher should know about Machine Learning El Mahdi El Mhamdi EPFL, Switzerland

Biography:

El Mahdi El Mhamdi is a PhD candidate at EPFL. His work is on the intersection between Artificial Intelligence and Distributed Algorithms. El Mahdi is interested in the robustness of machine learning abstractions such as neural networks or distributed settings for stochastic gradient descent, focusing on Byzantine fault tolerance and
safety. Before his PhD, El Mahdi worked as a research engineer in condensed matter physics and launched, with professor Rachid Guerraoui, the online learning project Wandida. He holds an engineering degree from the french École Polytechnique and a master of Science from the Swiss Federal Institute of Technology (EPFL).

Abstract:

Machine learning(ML) is the most active area in computer science research today. For the distributed computing community, ML poses a fertile set of problems and many opportunities of interaction
with our filed. In this tutorial, I will present the basics of machine learning needed for this interdisciplinary interaction to take place. Finally, I will illustrate this interaction with a recent work on Byzantine failures in neural networks.