Contact at email: kpb@caltech.edu
Office Contact: CMS Caltech
Pronouns: he/him
First name pronunciation: key-shun (Kishan)
My research interests in machine learning are motivated by addressing challenges for autonomous solutions to real-world problems. Specifically, I work on the theoretical foundations of reinforcement learning algorithms, tackling the simulation to real-world performance gaps by providing robust solutions. My future research will address such environmental uncertainties in more general domains, like multi-agent, imitation learning, human feedback, and more systems. In addition to statistical reinforcement learning, my daily drives and expertise span several areas like optimization, high-dimensional probability, multi-armed bandits, online learning, and stochastic theory.
I am a Postdoctoral Scholar Research Associate in Computing and Mathematical Sciences department in the Division of Engineering and Applied Science at California Institute of Technology (Caltech), hosted by Prof. Adam Wierman and Prof. Eric Mazumdar. I have been honored as a recipient of the PIMCO Postdoctoral Fellow in Data Science award. Before joining for the postdoc, I got my Ph.D. at Texas A&M University in August 2023, where I had the privilege of being advised by Prof. Dileep Kalathil.
Prior to joining the PhD program, I worked for Qualcomm Bengaluru in the Corporate Research and Development division. Before that, I completed my Masters in Communication and Networks from the Department of Electrical Communication Engineering at the Indian Institute of Science (IISc) under the guidance of Prof. Rajesh Sundaresan. It was during my master's program at IISc that I got introduced to the world of research. I am truly grateful to Rajesh for paving the way in this rich world called academic research. I completed my Bachelors from the Department of Electronics and Communication Engineering at PES Institute of Technology (now PES University) in India.
I am in the job market for full-time faculty positions and core industrial research positions starting in 2025. Please feel free to reach out! More details are under the Job Market tab.
Recent Career News
[March 2025] I am excited and grateful to receive the ML and Systems Rising Stars 2025 award! Awarded to the cohort of 38 among 150+ applications.
[March 2025] A theoretical framework bridging robust reinforcement learning and offline reinforcement learning via pessimism is to appear in L4DC 2025!
[February 2025] I am attending the Information Theory and Applications (ITA) workshop from 9-14 February at Bahia Resort, San Diego. I am grateful to be part of an amazing cohort of speakers at their Graduation Day scheduled for Feb 12.
[February 2025] At the IEOR Department and C-MInDS in IIT Bombay, I am giving a seminar on sequential learning algorithms making robust decisions against sim-2-real gaps.
[January 2025] A theoretical framework addressing transfer setup in reinforcement learning is to appear in AISTATS 2025! Led by my undergrad mentee! Accepted as oral (~2% of 583 accepted works).
[January 2025] New fundamental result on game-theoretic equilibria explaining human behaviors in multi-agent scenarios is appearing in ICLR 2025! Accepted as oral (1.8% of 11500 submitted works).
[January 2025] At Fujitsu Research Bengaluru and Tencent Seattle, I am giving tech talks on sequential learning algorithms making robust decisions against sim-2-real gaps.
[December 2024] I am attending the IndoML workshop from 21-23 December at BITS Pilani Goa Campus. I am grateful to be part of an amazing cohort of speakers at their Early Career Research Forum scheduled for Dec 23.
[December 2024] At Nokia Bengaluru and CSA Department IISc (video available here), I am giving a tutorial on "Basics of Reinforcement Learning: An Introduction." I am grateful to be part of an amazing cohort of speakers at the Theory CS Winter School, CSA, IISc, 2024.
[November/December 2024] At Microsoft Research Bengaluru, AI and CS Departments in IIT Hyderabad (video available here), and DSAI, EE, CS Departments in IIT Madras, I am giving tech talks on sequential learning algorithms making robust decisions against sim-2-real gaps.
[November 2024] I am serving as an Area Chair for AISTATS 2025.
[October 2024] I am co-chairing the "Strategic and Distributionally Robust Sequential Decision Making" session at the INFORMS Annual Meeting in Seattle, WA.