BEGIN:VCALENDAR
PRODID:-//eluceo/ical//2.0/EN
VERSION:2.0
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:www.tcs.tifr.res.in/event/1359
DTSTAMP:20231130T095622Z
SUMMARY:Towards Computation- and Communication-Efficient Distributed Learni
 ng
DESCRIPTION:Speaker: Pranay Sharma (Carnegie Mellon University\, Pittsburgh
 \, Pennsylvania)\n\nAbstract: \n\nModern machine learning (ML) systems rel
 y on the data collected at the edge devices to power diverse applications 
 like predictive typing\, personalized recommendations\, and real-time traf
 fic updates. However\, data privacy concerns and network bandwidth constra
 ints preclude gathering the entire dataset at a central location for furth
 er processing. In the past few years\, federated learning (FL) has emerged
  as a natural solution to this problem. The edge devices in FL maintain ex
 clusive control of their data and in return\, shoulder part of the computa
 tional load of the central server. Google and Apple have already deployed 
 FL to improve GBoard and Siri.\n \nIn this talk\, I will discuss my work 
 addressing several challenges in FL. Despite extensive research over the p
 ast few years\, the underlying optimization problems solved by most work a
 re simple minimization. However\, many ML applications\, like GANs\, robus
 t learning\, and reinforcement learning\, can be modeled as min-max proble
 ms. I will first describe my work solving nonconvex min-max problems in a 
 federated setting. In addition to achieving state-of-the-art theoretical c
 omputation-communication guarantees\, this work interestingly even improve
 s the existing centralized methods. Next\, I will also talk about my work 
 on FL systems solving minimization problems\, where I quantify the impact 
 of limited device participation\, where only a small fraction of all the d
 evices may be available at any time. I will then discuss a reinforcement l
 earning problem in a federated setting\, where we prove linear speedup in 
 the presence of Markov noise\, answering an existing open question. Finall
 y\, I will conclude with some future directions I'm excited about and my b
 roader research vision.\n
URL:https://www.tcs.tifr.res.in/web/events/1359
DTSTART;TZID=Asia/Kolkata:20231204T110000
DTEND;TZID=Asia/Kolkata:20231204T120000
LOCATION:via Zoom in A201
END:VEVENT
END:VCALENDAR