Hi! I am a researcher interested in problems related to algorithmic fairness, data and tech ethics, and community-focused AI development. Broadly, my work studies the interactions between society and automation and addresses various questions around the potentially harmful impacts of AI applications.
I am currently a Postdoctoral Associate, working with Walter Sinnott-Armstrong and Jana Schaich Borg at Duke University, and Hoda Heidari and Vincent Conitzer at Carnegie Mellon University. I am based at the Duke Philosophy department and also affiliated with the Kenan Institute of Ethics.
I received my Ph.D. from the Department of Statistics and Data Science at Yale University, where I was advised by Elisa Celis. While at Yale, I was also a Resident Fellow at the Information Society Project (ISP) at Yale Law School and a 2022 Policy Fellow at the Yale Institute for Social and Policy Studies (ISPS). Prior to this, I spent two years at École Polytechnique Fédérale de Lausanne (EPFL). Even before that, I completed my Bachelor’s and Master’s degrees in Computer Science and Engineering at the Indian Institute of Technology, Kanpur under the supervision of Rajat Mittal.
Vijay Keswani, L. Elisa Celis
Addressing Strategic Manipulation Disparities in Fair Classification
EAAMO 2023. arxiv
Vijay Keswani
Social Media Platform Structures and Their Implications
EcAI Workshop 2023. pdf
Vijay Keswani, L. Elisa Celis, Matthew Lease, Krishnaram Kenthapadi
Designing Closed-Loop Models for Task Allocation
HHAI 2023. arxiv
Vijay Keswani, L. Elisa Celis
An Anti-Subordination Approach to Fair Classification
EAAMO 2022 (Non-archival). ssrn
Florian Evéquoz, Johan Rochel, Vijay Keswani, L. Elisa Celis
Diverse Representation via Computational Participatory Elections – Lessons from a Case Study
EAAMO 2022. arxiv
Vijay Keswani, Oren Mangoubi, Sushant Sachdeva, Nisheeth K. Vishnoi
A Convergent and Dimension-Independent Min-Max Optimization Algorithm
ICML 2022. arxiv
Vijay Keswani, Matthew Lease, Krishnaram Kenthapadi
Designing human-in-the-loop approaches for closed deferral pipelines
BHCC 2021 (Non-archival). arxiv
Vijay Keswani, L. Elisa Celis
Auditing for Diversity using Representative Examples
ACM SIGKDD 2021. arxiv
Vijay Keswani, Matthew Lease, Krishnaram Kenthapadi
Towards Unbiased and Accurate Deferral to Multiple Experts
AIES 2021. arxiv
L. Elisa Celis, Lingxiao Huang, Vijay Keswani, Nisheeth K. Vishnoi
Fair Classification with Noisy Protected Attributes: A Framework with Provable Guarantees
ICML 2021. arxiv
Vijay Keswani, L. Elisa Celis
Dialect Diversity in Text Summarization on Twitter
The Web Conference 2021. arxiv
L. Elisa Celis, Vijay Keswani
Implicit Diversity in Image Summarization
CSCW 2020. arxiv
L. Elisa Celis, Vijay Keswani, Nisheeth K. Vishnoi
Data preprocessing to mitigate bias: A maximum entropy based approach
ICML 2020. arxiv
L. Elisa Celis, Vijay Keswani
Improved Adversarial Learning for Fair Classification
arxiv
L. Elisa Celis, Sayash Kapoor, Farnood Salehi, Vijay Keswani, Nisheeth K. Vishnoi
A Dashboard for Controlling Polarization in Personalization
Invited paper in AI Communications 2019
(Previously in IJCAI-ECAI 2018)
L. Elisa Celis, Lingxiao Huang, Vijay Keswani, Nisheeth K. Vishnoi
Classification with Fairness Constraints: A Meta-Algorithm with Provable Guarantees
ACM-FAT* 2019. arxiv
L. Elisa Celis, Vijay Keswani, Damian Straszak, Amit Deshpande, Tarun Kathuria, Nisheeth K. Vishnoi
Fair and Diverse DPP-based Data Summarization
ICML 2018. arxiv