Multi-armed Bandits for Wireless Networks: Introduction, State-of-the-art Works and Potential Research Directions
Multi-armed bandits (MAB) are popular online machine learning algorithms widely used for real-time decision-making without the need for prior training. They have been widely used for online advertisement selection to increase the number of clicks, clinical trials to identify best drugs, news personalization, decision making in financial markets and resource selection in wireless networks, internet of things (IoT) and robotics. In this talk, we will focus on introduction to MAB algorithms and their applications in wireless networks. We will also discuss current state-of-the-art works from algorithms as well as architecture perspectives followed by potential research directions.
Dr. Sumit J Darak received UG degree from Pune University, Pune, India, in 2007, and the Ph.D. degree from Nanyang Technological University (NTU), Singapore, in 2013. He is currently an Associate Professor with the Indraprastha Institute of Information Technology, Delhi (IIIT-Delhi), Delhi, India. His research team is the recipient of the “DST Inspire Faculty Award”, "Best Thesis Award" in COMSNETs 2022, “Design Contest Winner'' VLSID 2022, “Best Paper Award” AIMLSystems 2021, “Second Best Poster Award” at COMSNETS 2019, “Second Best Paper Award” at DASC 2017, “Best Demo Award” at CROWNCOM 2016, “Young Scientist Paper Award” at URSI 2014 and 2017, and “Best Paper Award” at NCSIPA 2009. He has received funding from MeiTy, DST, National Instruments (NI) and VVDN Technologies. His current research interests are intelligent and reconfigurable architectures for applications such as wireless communications, signal processing, artificial intelligence, etc.