Seminar: Learnings from industrial research on privacy and machine learning on wireless networks.

Speaker: Ilias Leontiadis

Date/Time: 21-Nov-2019, 16:00 UTC

Venue: Roberts 421

Details

Abstract:

Mobile ISPs operate a vast country-wide infrastructure that is comprised of thousands of cellular sites. These sites are also a unique vantage point where we can better understand the privacy and security implications of the users’ actions. In this talk we will present some of the recent research work that has been conducted towards this direction.  More specifically, we will discuss about the privacy implications of location sharing and how this can be used to deanonymize users in a country-wide scale. Furthermore, we will look into how network traces can help us identify and stop illicit cryptocurrency mining on the Web. We will also explore the landscape of mobile browsers and how they relate to user privacy.

Furthermore, the cellular infrastructure offers great opportunities to accelerate  computation. The second part of this talk will focus on more recent work conducted while at samsung AI. We will examine recent developments on accelerating on-device AI with the help of edge computing and the optimizations required for on-device computation.

Bio:

Ilias Leontiadis is currently a Senior Research Scientist at Samsung AI Cambridge. Before, he was as Senior Researcher at Telefonica Research and a research fellow at the University of Cambridge.

His research interests include mobile systems, machine learning and wireless networks. He is leading the efforts to enable edge and cloud offloading of complex AI tasks from resource-constrained devices such as wearables and IoT.  Previously he was working on machine learning for cellular networks. Apart from the industrial research, he has been involved in multiple academic conferences (WWW, Mobisys, MobiCom, CoNext, ICWSM, CSCW, IMC, SIGCOMM, etc).

leontiadis.net

Add to Calendar

This page was last modified on 27 Mar 2014.