BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Optimization algorithms for heterogeneous federated learning
DTSTART:20211207T170000
DTEND:20211207T180000
DTSTAMP:20260407T034936Z
UID:a898c243a06b578a5884dbc83a5a73c618fe3ce008b1c1e113d17e7e
CATEGORIES:Conferences - Seminars
DESCRIPTION:\n\n\nTitle for the talk: \n\nOptimization algorithms for het
 erogeneous federated learning\n\n \n\nAbstract for the talk:\nA traditio
 nal machine learning pipeline involves collecting massive amounts of data 
 centrally on a server and training models to fit the data. However\, incre
 asing concerns about the privacy and security of user's data\, combined wi
 th the sheer growth in the data sizes has incentivized looking beyond such
  traditional centralized approaches. Federated learning proposes instead f
 or a network of data holders to collaborate together to train models witho
 ut transmitting any data. This new paradigm minimizes data exposure\, but 
 inherently faces some fundamental optimization challenges posed by non iid
  data across the users' data. We will discuss our understanding\, and prog
 ress in tackling these problems.\n\n \n\nPapers covered:\n\n1. SCAFFOLD: 
 Stochastic Controlled Averaging for Federated Learning https://arxiv.org/
 abs/1910.06378\n\n2. Mime: Mimicking Centralized Stochastic Algorithms in 
 Federated Learning. https://arxiv.org/abs/2008.03606\n\n \n\nBio: \n\nD
 r. Praneeth Karimireddy recently finished his PhD at EPFL advised by Prof
 . Martin Jaggi. His main research interest is developing intelligence inf
 rastructure for collaborative learning. His research has been awarded wit
 h a Dimitris N. Chorafas Foundation Prize\, and a best paper award at FL-I
 CML 2021. Supported by an SNSF fellowship\, he will be joining as a postd
 oc with Mike Jordan's group at UC Berkeley in Spring 2022.\n\n\n
LOCATION:https://epfl.zoom.us/j/68198575602
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR