An introduction to fairness in reinforcement learning.

Speaker:
Organiser:
Umang Bhaskar
Date:
Thursday, 5 Sep 2024, 14:00 to 15:00
Venue:
A-238
Category:
Abstract

In a series of three talks, we present an overview of work in bandit optimization, including stochastic and adversarial bandits, and will focus on variants of multi-armed bandits where fairness is an additional constraint. In Fair-MAB, for example, in addition to the objective of maximizing the sum of expected rewards, the algorithm also needs to ensure that at any time, each arm is pulled at least a pre-specified fraction of times. We investigate the interplay between learning and fairness in such settings, and also investigate the cost of fairness.

Short Bio:

Vishakha Patil works in online learning and optimization, game theory, and mechanism design. She completed her PhD from IISc, supported by a Google Fellowship and a CII-SERB PMRF. She has also been a Heidelberg Laureate Forum Young Researcher, and her MTech thesis received an Honourable Mention for the Best Thesis Award at CSA, IISc.