Hassan Mortagy
I am currently an OR Scientist at Roadie, developing algorithms for optimal routing and dynamic pricing. I recieved my Ph.D. in Operations Research from the H. Milton Stewart School of Industrial & Systems Engineering at Georgia Tech in May 2023, minoring in Machine Learning. I was advised by Prof. Swati Gupta, and my research aimed to exploit (combinatorial) structures present in Optimization and Machine Learning problems to develop novel and efficient algorithms with provable theoretical guarantees.
I interned as an Applied Scientist at Amazon during the Summer of 2022. Before this, I obtained my M.S. In Management Science and Engineering from Columbia University. During my time at Columbia, I also worked as a part-time analytics consultant for several companies. Prior to that, I obtained my B.S. in Construction Engineering from the American University in Cairo, where my focus was on project management.
Research Interests:
Convex and Combinatorial Optimization
Machine Learning
Revenue Management
Latest News:
- May 2023. I joined Roadie as an OR Scientist
- January 2023. I gave a talk on Exploiting Combinatorial Structures in Machine Learning and Convex Optimization at Cornell ORIE
- November 2022. I received the Shabbir Ahmed Optimization Research Excellence Award.
- October 2022. I presented our work on Projections over Submodular Polytopes at INFORMS [Slides]
- September 2022. I will be presenting a poster on Projections over Submodular Polytopes at the 2022 Cornell Young Researchers Workshop.
- May 2022. We got an honorable mention for best poster award on Faster Iterative Projections on Submodular Polytopes at the Mixed Integer Programming (MIP) workshop, 2022. [photo of award]
- April 2022. I will be interning at Amazon research as an Applied Scientist during the summer of 2022.
- March 2022. Our work on projections over Submodular polytopes has been selected as a finalist for the MIP Workshop 2022 poster competition
- December 2021. I presented a poster on our "Faster Iterative Projections" paper at NeurIPS 2021. [Poster]
- October 2021. I presented "Electrical Flows over Spanning Trees" at INFORMS [Video]
- September 2021. Our paper on "Reusing Combinatorial Structure: Faster Projections over Submodular Base Polytopes" has been accepted to NeurIPS 2021. [paper link]
- June 2021. We developed a toolkit for speeding up iterative projections over submodular polytopes. Projection-based optimization algorithms enjoy near-optimal convergence rates, but suffer from a computational bottleneck of computing "projections'' in every iteration. On the other hand, conditional gradient variants are efficient but result in suboptimal rates. Our paper is motivated by this trade-off in runtime v/s convergence rates. [paper link]
- April 2021. I am organizing a session on Bridging Discrete and Continuous Optimization at INFORMS 2021.
- January 2021. "Electrical Flows over Spanning Trees" got accepted into Mathematical Programming B for publication.
- December 2020. I presented a poster on our "Walking in the Shadow" paper at NeurIPS 2020. [Poster]
- November 2020. I Received the Algorithms and Randomness Center (ARC) TRIAD research fellowship for Spring 2021 semester.
- November 2020. Invited talk and session at INFORMS 2020 annual meeting. [video]
- October 2020. I gave a talk on our paper “Walking in the Shadow: A New Perspective on Descent Directions for Constrained Minimization” in the ISyE student seminar.
- September 2020. Our paper “Walking in the Shadow: A New Perspective on Descent Directions for Constrained Minimization” got accepted into NeurIPS 2020. This is joint work with Swati Gupta and Sebastian Pokutta. [paper link] [video summary]
CONTACT:
e-mail: hmortagy@gatech.edu