Recent Work

We study the problem of learning personalized decision policies from observational data while accounting for possible unobserved confounding in the data-generating process.

We study the downstream impacts of systematically censored dataset construction on attempts to adjust for fairness. Under certain conditions, fairness-adjusted dclassifiers induce residual unfairness that perpetuates the same injustices that biased the data to begin with.
ICML 2018,2018

Presented at AISTATS 2018. Finalist, INFORMS Data-Mining and Decision Analytics Workshop,2017


. Sequential Decision Making over Networks: Coupon Targeting. ICML Workshop on Data Efficient Machine Learning), 2016.

. Dynamic Task Allocation for Crowdsourcing Settings . ICML Workshop on Data-Efficient Machine Learning, 2016.


. Analysis of the Spatial Distribution of Semantic Meaning using Tweets from Manhattan. Interface Symposium for Data Science, 2015.

Recent & Upcoming Talks

Recent Posts

keywords: civic data/open data, applied microeconomics, graphic design, photography.