Akshit Kumar

I am third year PhD student in the Decision, Risk and Operations division at Columbia Business School , where I am fortunate to be advised by Prof. Yash Kanoria and Prof. Omar Besbes . My current research interests are network revenue management, dynamic matching and reinforcement learning theory.

Previously, I completed my Masters in Electrical Engineering from University of Michigan where I worked with Prof. Vijay Subramanian on theoretical reinforcement learning and stochastic control. I have been fortunate to work with an amazing set of advisors in the past. I completed my bachelors thesis with Prof. Rahul Vaze at Tata Institute of Fundamental Research . I also interned at Microsoft Research India, where I was advised by Dr. Ranveer Chandra and Dr. Manohar Swaminathan .

I graduated with a B.Tech (Hons.) in Electrical Engineering from Indian Institute of Technology, Madras .

Publications

The Multi-secretary Problem with Many Types
Omar Besbes, Yash Kanoria, Akshit Kumar
EC'22 | ACM Conference on Economics and Computation
pdf| abstract| cite

Low-cost aerial imaging for small holder farmers
Aditya Jain, Zerina Kapetanovic, Akshit Kumar, Vasuki Narasimha Swamy, Rohit Patil, Deepak Vasisht, Rahul Sharma, Manohar Swaminathan, Ranveer Chandra, Anirudh Badam, Gireeja Ranade, Sudipta Sinha, Akshay Uttama Nambi SN
COMPASS'19 | ACM SIGCAS Conference on Computing and Sustainable Societies
pdf| abstract| cite

Speed scaling under QoS constraints with finite buffer
Parikshit Hegde, Akshit Kumar, Rahul Vaze
WiOpt'18 | International Symposium on Modeling and Optimization in Mobile, Ad hoc, and Wireless Networks
pdf| abstract| cite

Working Papers

Feature-Based Dynamic Matching
Yilun Chen, Yash Kanoria, Akshit Kumar
Working Paper
pdf

Finite Time Analysis of Empirical Dynamic Programs
Akshit Kumar, Daniel Vial, Vijay Subramanian
Under Preparation

Breaking the Unit Throughput Barrier in Distributed Systems
Akshit Kumar, Parikshit Hegde, Rahul Vaze, Amira Alloum, C├ędric Adjih
Working Paper
pdf| abstract| cite

Talks

The Multi-secretary Problem with Many Types
INFORMS Annual Meeting, Indianapolis, Indiana, October 2022
EC'22, Boulder, Colorado, July 2022
MSOM Annual Conference, Munich, Germany, June 2022
RMP Annual Conference, Online, June 2022
flash talk|video| poster

Feature Based Dynamic Matching
INFORMS Annual Meeting, Indianapolis, Indiana, October 2022

IIT Madras
2014 - 2018
Microsoft Research India
Summer 2016
STCS, TIFR
Summer 2017
Nokia Bell Labs
Summer 2018
University of Michigan
2018 - 2020
Columbia University
2020 - Present