Seminars

View all Seminars  |  Download ICal for this event

Provable and Efficient Algorithms for Federated, Reinforcement and Batch Learning

Series: Department Seminar

Speaker: Dr. Avishek Ghosh, Univ. of California, San Diego, USA

Date/Time: Dec 16 08:00:00

Location: Online Meeting - ON-LINE

Faculty Advisor:

Abstract:
In this talk, I will propose and analyze iterative algorithms that are computationally efficient, statistically sound and adaptive to the problem complexity. Three different frameworks in which data is presented to the learner are considered : (a) Federated (Distributed) Learning (FL) setup, where data is only available at the edge, and a center machine learns various models via iteratively interacting with the edge nodes; (b) Reinforcement Learning (RL) framework, where agents successively interacts with the environment to learn an optimal policy and (c) Supervised batch setup, where all the data and label pairs are available to the learner at the beginning. In particular, I will explain my contribution in the FL and RL framework and (very) briefly mention the supervised batch setup.
In the Federated Learning (FL) framework, I will address the canonical problems of device heterogeneity, communication bottleneck and adversarial robustness for large scale high dimensional problems, and propose efficient and provable algorithms that obtain the optimal statistical rate. In The Reinforcement Learning (RL) setup, I will focus on the problem of structure adaptation (model selection) for large scale problems---an aspect most practical RL based systems like autonomous cars, recommendation systems crucially require. The model selection problem is studied both with function approximation (parameterized RL) as well as for the generic model based RL setup, and the algorithms we propose obtain (near) optimal regret guarantees.
Teams Meeting Link: https://teams.microsoft.com/l/meetup-join/19%3ameeting_MGUwNjY3NmEtZWZkYy00NzJkLWIzZjMtOWUzMTAyOTgwNzg3%40thread.v2/0?context=%7b%22Tid%22%3a%226f15cd97-f6a7-41e3-b2c5-ad4193976476%22%2c%22Oid%22%3a%224bcd3d56-e405-4b06-99fb-27742262f261%22%7d

Speaker Bio:
I am an HDSI (Data Science) Post-doctoral fellow at the University of California, San Diego, working with Prof. Arya Mazumdar and Prof. Tara Javidi. Prior to this, I completed my PhD from the EECS department of UC Berkeley, advised by Prof. Kannan Ramchandran and Prof. Aditya Guntuboyina. My research interests are broadly in Theoretical Machine Learning, including Federated Learning and multi-agent Reinforcement/Bandit Learning. I spent the summer of 2020 working as a Research Scientist at Amazon, New York; in the Supply Chain Optimization Team (SCOT), where I worked with Dean Foster and Alexandar (Sasha) Rakhlin. Before coming to Berkeley, I completed my masters degree from Indian Institute of Science (IISc), Bangalore working with Prof. Anurag Kumar, and Prof. Aditya Gopalan. Before IISc, I spent 4 years at Jadavpur University, where I completed my bachelors degree and worked with Prof. Amit Konar. I am a recipient of the Excellence Award from UC Berkeley and Gold medal from Jadavpur University.

Host Faculty: R. Govindarajan