Amel O. Awadelkarim

I’m a PhD student at Stanford University at the Institute for Computational and Mathematical Engineering (ICME), advised by Johan Ugander. I’m interested in applying tools from statistics, network science, causal inference, and machine learning to study people and the systems we all participate in. My current research uses discrete choice models to model household preferences for school choice, in partnership with the San Francisco Unified School District, for the purpose of inference and policy evaluation. In the past, I’ve done work in recommender systems and balanced graph partitioning.

I’m also very good at ultimate frisbee, earning World and National titles in 2018 and 2021, am always listening to music, and drink a cup of homemade chai (almost) everyday. Welcome to my website!

news

Oct 16, 2023	Began research scientist internship at Meta in Central Applied Science on the Graph Science and Statistics team.
Oct 13, 2023	Gave my ~spooky~ Friday the 13th thesis defense, entitled Statistical Modeling of Choices and Rankings.
Aug 6, 2023	Published school choice preference modeling work in proceedings of KDD 2023, and gave oral and poster presentations at the conference in Long Beach, CA.

selected publications

Rank-heterogeneous preference models for school choice

Amel Awadelkarim, Arjun Seshadri, Itai Ashlagi, and 2 more authors

In Proceedings of the 29th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2023

Abs arXiv Code

School choice mechanism designers use discrete choice models to understand and predict families’ preferences. The most widely-used choice model, the multinomial logit (MNL), is linear in school and/or household attributes. While the model is simple and interpretable, it assumes the ranked preference lists arise from a choice process that is uniform throughout the ranking, from top to bottom. In this work, we introduce two strategies for rank-heterogeneous choice modeling tailored for school choice. First, we adapt a context-dependent random utility model (CDM), considering down-rank choices as occurring in the context of earlier up-rank choices. Second, we consider stratifying the choice modeling by rank, regularizing rank-adjacent models towards one another when appropriate. Using data on household preferences from the San Francisco Unified School District (SFUSD) across multiple years, we show that the contextual models considerably improve our out-of-sample evaluation metrics across all rank positions over the non-contextual models in the literature. Meanwhile, stratifying the model by rank can yield more accurate first-choice predictions while down-rank predictions are relatively unimproved. These models provide performance upgrades that school choice researchers can adopt to improve predictions and counterfactual analyses.