I am a final-year Operations Research Ph.D. candidate at Stanford and am very lucky to work with Professors Susan Athey and Emma Brunskill. Before coming to Stanford, I studied Mathematics and Computer Science at Chennai Mathematical Institute and was fortunate to be mentored by Professor Siddharth Barman. In 2022, I spent a wonderful summer interning as an applied scientist at Amazon, working on offline RL with Branislav Kveton.

I am broadly interested in developing algorithmic and data-driven tools for decision-making, especially for systems that adapt to human preferences/responses. My recent work has focused on addressing the practical shortcomings of existing contextual bandit algorithms to make them reliably outperform non-adaptive randomized control trials. Over the course of my Ph.D., I have been developing a modular and reliable contextual bandit pipeline. My work has led to novel technical insights on personalized decision-making and uncertainty quantification that are of broader interest. In the past, I have also worked on algorithmic game theory.

Email: [myfirstname]sk[at]stanford[dot]edu 

Links to my: Google scholar, LinkedIn.

Selected Pre-Prints and Publications: (Complete List)

• Towards Costless Model Selection in Contextual Bandits: A Bias-Variance Perspective [arXiv] 

Sanath Kumar Krishnamurthy*, Adrienne Propp*, Susan Athey.

AISTATS 2024.

[First approach to incorporate bias-variance trade-off into model selection for contextual bandits! Demonstrated the feasibility of costless model selection in contextual bandits under fairly benign conditions! Developed a novel misspecification test to evaluate prediction error bounds! The misspecification test enabled assumption-free guarantees in  [NeurIPS 2023].]

• Proportional Response: Contextual Bandits for Simple and Cumulative Regret Minimization [arXiv] 

Sanath Kumar Krishnamurthy, Ruohan Zhan, Susan Athey, Emma Brunskill.

NeurIPS 2023.

[New form of uncertainty for contextual decision making! First general-purpose regression-based algorithm for simple regret minimization! Also, the first general-purpose algorithm that flexibly trades off simple and cumulative regret guarantees in contextual bandits!]

• Flexible and Efficient Contextual Bandits with Heterogeneous Treatment Effect Oracles [arXiv]

Aldo Gael Carranza*,  Sanath Kumar Krishnamurthy*, Susan Athey.

AISTATS 2023.

[Demonstrate the theoretical and empirical benefits of using heterogeneous treatment effect estimation in contextual bandits! Uncovered favorable properties of using R-learner for downstream decision-making tasks!]

• Contextual Bandits in a Survey Experiment on Charitable Giving: Within-Experiment Outcomes versus Policy Learning [arXiv]

Susan Athey*, Undral Byambadalai*, Vitor Hadad*, Sanath Kumar Krishnamurthy*, Weiwen Leung*, Joseph Jay Williams*

[Took contextual bandits to the real world (charitable giving)!]

• Finding Fair and Efficient Allocations [arXiv]

Siddharth Barman*, Sanath Kumar Krishnamurthy*, Rohit Vaish*

EC, 2018.

[Developed the first pseudo-polynomial time algorithm for computing fair (EF1) and Pareto optimal (PO) allocations of indivisible goods among agents with additive valuations! This is technically fascinating because even checking if an allocation is PO is co-NP complete. Our techniques to search over such allocations were used in several subsequent papers.]