Kesav Kaza is a postdoctoral researcher in human-automation collaboration working under the supervision of Prof. Jerome Le Ny and Prof. Aditya Mahajan. His interests include decision making under uncertainty, human in the loop systems, sequential decision analysis and reinforcement learning. He received PhD in Electrical Engineering in Feb. 2020 from Indian Institute of technology Bombay, Mumbai, India, under the supervision of Prof. Shabbir Merchant. His thesis was Sequential decision making under uncertainty with limited observation capability : Application to wireless networks.
His current work is on designing algorithms for decision referrals in human-automation teams. This research considers a scenario where an automation takes a look at classification tasks and decides to refer a subset of the tasks to a human. The optimal referral algorithm needs to consider its uncertainty about the tasks along with human factors such as workload which impact human classification performance. Another aspect of this problem includes devising algorithms to learn from human feedback.