Abdellah Aznag 
Postdoctoral Researcher - Harvard Business School AI Institute

I am currently a Postdoctoral Researcher at the Harvard Business School AI Institute. 

My research aims at developing the foundational principles to strategically collect data within complex systems. The value of a final decision is inextricably linked to the quality of the data that informs it. Yet, data collection is often treated as a passive or ad-hoc process, which leads to critical oversight, biased conclusions, and significant opportunity costs. This reality motivates the central question of my work: How do we design policies that optimally balance the cost of acquiring new information against the value it provides for a decision? 

I received my PhD from the Department of Industrial Engineering and Operations Research at Columbia University, where I was advised by Prof. Rachel Cummings and Prof. Adam N. Elmachtoub. Prior to joining Columbia, I earned a BS and MS in Applied Mathematics from Ecole Polytechnique.  I grew up in El Jadida, Morocco.

Papers

Survey Design for Fair Resource Allocation with Max Biggs, Ferdinando Fioretto and Joonhyuk Ko

1. • Submitted

Sample Complexity of Peer Prediction with Robin Bowers, Rachel Cummings, Jason Hartline, Matthew vonAllmen and Bo Waggoner

1. • Preliminary version presented at EC26

Designing Lower Bounds for Active Learning in Multi-Armed Bandits with Rachel Cummings and  Adam N. Elmachtoub

1. • Preliminary version presented at COLT26

An Active Learning Framework For Multi-Group Mean Estimation, with Rachel Cummings and Adam N. Elmachtoub

1. • Preliminary version appeared in the proceedings of NeurIPS23

   • Major revision in Management Science

   • Finalist, INFORMS George Nicholson Student Paper Competition, 2026

MNL-Bandits with knapsacks: a near-optimal algorithm, with Vineet Goyal and Noemie Perivier

1. • Preliminary version presented at EC21

   • Minor revision in Operations Research

Technical reports

The tropicalization of the entropic barrier, with Xavier Allamigeon, Stephane Gaubert, and Yassine Hamdi

Teaching

Teaching Assistant

• IEOR4004: Optimization (Fall 2024)

• IEOR3658: Probability for Engineers (Fall 2023, Spring 2024)

• IEOR4507: Healthcare Operations Management (Spring 2023)

• IEOR4106: Stochastic Modeling (Fall 2021)

• IEOR4650: Business analytics (Spring 2021)

Instructor

• Columbia Science Honors Program: Introduction to algorithms (Fall 2021, Fall 2022)