Vashist Avadhanula

PhD Candidate Decision, Risk and Operations Graduate School of Business, Columbia University Office: Uris Hall-4C Email: vavadhanula18 At gsb Dot columbia Dot edu My CV I am on the academic job market this year.

Brief Biography

I am a graduate student in Decision, Risk & Operations division of Columbia Business School working with Prof. Assaf Zeevi, Prof. Vineet Goyal and Prof. Shipra Agrawal in the intersectional areas of machine learning and operations management. My current work focuses on dynamic learning in revenue management under demand uncertainty. I'm also interested in data driven modeling and estimation. I spent the summer of 2016 as a research intern in Flipkart, India's largest online retail store, where I developed optimization models to improve the average delivery times. I'm also working with their Search and Discover team to develop a framework to understand user substitution behavior and improve product recommendations.

Prior to joining Columbia, I was a dual degree student in Electrical Engineering department at Indian Institute of Technology Bombay (IIT-B), where I spent an year in the Computer Science & Engineering department working with Prof. Sunita Sarawagi on my thesis in the area of probabilistic graphical models. I graduated from IIT-B in 2013 with Bachelor's in Electrical Engineering and Master's in Communication & Signal Processing.

Dissertation

Publications

Thompson Sampling for the MNL-Bandit In Conference on Learning Theory (COLT), 2017.
(joint with Shipra Agrawal, Vineet Goyal and Assaf Zeevi )
Journal version in Preparation.

MNL-Bandit: A Dynamic Learning Approach to Assortment Selection Submitted to Operations Research.
(joint with Shipra Agrawal, Vineet Goyal and Assaf Zeevi )

A near-optimal algorithm for exploration-exploitation approach for assortment selection. In EC, 2016.
(joint with Shipra Agrawal, Vineet Goyal and Assaf Zeevi )

On the tightness of an LP relaxation for Rational Optimization. Operations Research Letters, 2016
(joint with Jalaj Bhandari, Vineet Goyal and Assaf Zeevi)

A few good predictions: Selective node labeling in a social network. In WSDM, 2014.
(joint with Gaurish Chaudhari and Prof. Sunita Sarawagi. )