Abdulrahman Ahmed
Abdulrahman is a recent Ph.D. graduate from the Department of Industrial Engineering at the University of Pittsburgh. His current research focuses on devising efficient techniques for estimating treatment effects from large-scale simulations (i.e., simulation metamodel), particularly in scenarios where sampling from these simulations proves to be resource-intensive. Previously, he worked on developing reinforcement learning methods for a financial problem called optimal execution problem, where the decision maker needs to devise a cost-saving strategy to maximize the gain from the execution. He got his MSc and BSc in operations research and computer science from Cairo University.
Google Scholar Link : https://scholar.google.com/citations?hl=en&authuser=1&user=kf8by4MAAAAJ
Research Interest
Metamodeling, Simulation Optimization, Large-scale Simulation, and Machine Learning.
Career Update
Data Scientist at Citizens Bank
Dissertation
Computationally Efficient Simulation-Based Policy Evaluation for Epidemic Interventions: Metamodeling, Sequential Design, and Localized Effect Estimation [link]
Summary: Abdulrahman Ahmed developed his dissertation towards a vision of precision public health with policy evaluation methods that are local, computationally tractable, and explicitly designed for heterogeneous populations. To that end, he proposed bi-level methods that combine Gaussian processes and regression models to efficiently learn correlations across US counties and between intervention levels and components when estimating county-level treatment effects, along with additional methods to optimize the sequential sampling of counties and the selection of regression models for interventions.
Publications
- "Optimized Model Selection for Estimating Treatment Effects from Costly Simulations of the US Opioid Epidemic", with M. A. Rahimian, and Mark S. Roberts. To be presented in the 2024 Annual Simulation Conference (ANNSIM'24). [arXiv:2403.15755][code]
- "Inferring Epidemic Dynamics Using Gaussian Process Emulation of Agent-Based Simulations", with M. A. Rahimian, and Mark S. Roberts. Winter Simulation Conference (WSC), December 10-13, 2023, San Antonio, TX. [IEEE:10408157][arXiv:2307.12186][code]
- "Estimating Treatment Effects Using Costly Simulation Samples from a Population-Scale Model of Opioid Use Disorder", with M. A. Rahimian, and Mark S. Roberts. IEEE International Conference on Biomedical and Health Informatics (BHI), October 15-18, 2023, Pittsburgh, PA. [IEEE:10313496][arXiv:2308.13040][code]
- "Optimizing Stock Market Execution Costs Using Reinforcement Learning", with A. Ghoneim, and M. Saleh. 2020 IEEE Symposium Series on Computational Intelligence (SSCI), December 1-4, 2020, Canberra, Australia. [IEEE:9308153]
- Pre-Print: "Computationally Efficient Estimation of Localized Treatment Effects in High-Dimensional Design Spaces using Gaussian Process Regression" [link]
Working Papers
County-Level Heterogeneity in Opioid Harm Reduction and Treatment Effects: A Calibrated Markov Model [medrxiv.org]
