Subhojyoti Mukherjee

330 North Orchard Street, Room 4110
Madison WI 53715
Preferred pronouns: he/him/his
Joined WID: 2019
smukherjee27@wisc.edu
https://subhojyoti.github.io/


Multi-Armed Bandits and its various applications in real-life scenarios

About

Subhojyoti is a PhD candidate at the University of Wisconsin–Madison from Fall, 2019. Previously, he was an MS (Research) scholar in the Computer Science and Engineering Department, IIT Madras from January 2015 to July 2018. He was advised by Dr. Balaraman Ravindran (CSE Department, IIT Madras) and Dr. Nandan Sudarsanam (Department of Management Studies, IIT Madras). Subhojyoti was associated with the RISE lab at IIT Madras and completed his BTech from Meghnad Saha Institute of Technology, Kolkata under the West Bengal University of Technology in 2013.

Education

  • PhD, Electrical and Computer Engineering, University of Wisconsin–Madison (in process)
  • MS, Computer Science, Indian Institute of Technology Madras
  • BTech, Computer Science, West Bengal University of Technology

Research Description

My research interests span the areas of Machine Learning, Reinforcement Learning, Online Optimization, and Recommender Systems. I work in the area of stochastic and non-stochastic Multi-armed Bandits (MAB). MABs find natural application in a lot of real-world scenarios where interactive learning is present between users and system. In recent years, MABs have been increasingly gaining attention in a number of inter-disciplinary areas such as online education, online medical/health recommendation, online advertising, worker productivity management, and several other interesting industrial applications. I focus on theoretical guarantees of MAB policies so that better policies with safer outcomes can be guaranteed with high probability.