ECE + Math at Duke

Zain Radwan

COMAP Scholar · Research Assistant · Captain of Duke Rugby

I work on machine learning, optimization, and quantitative research.

ResumeGitHubLinkedInEmail
Zain Radwan

At a glance

Research, modeling, and markets

Featured

1 of 6 worldwide

COMAP Scholar

Outstanding Winner — MCM/ICM 2026

One of 30 Outstanding Winner teams from 93,977 participants across 28 countries — then selected as one of 6 COMAP Scholars globally. Built statistical and optimization models to evaluate strategic decisions for Crystal Palace FC under nonlinear financial risk.

MAA AwardLeonhard Euler AwardUMAP Journal (under review)

Read about the paper

Research

Duke Cosmology & Astrophysics Research Center

Satellite collision modeling for low Earth orbit — Monte Carlo simulation over ~40,000 tracked objects and QUBO optimization for satellite routing.

Monte CarloQUBOOrbital mechanics

AI / ML

Tree Crown Detection

CNN and transformer-based computer vision models estimating tree crown cover from PlanetScope satellite imagery for Duke's Ecology Department.

CNNsTransformersPlanetScope

Experience

  • IMC Launchpad

    Quantitative trading track — 1 of 60 selected. Simulated markets, expected value, risk management.

  • Duke Research

    Cosmology (satellite risk) and Ecology (geospatial ML) labs, plus teaching assistant for Duke Mathematics.

Leadership

Captain, Duke Rugby

Leading training, strategy, and culture for a Division 3 collegiate rugby program.

Contact

Email GitHub LinkedIn

Skills

PythonMachine LearningOptimizationStatisticsComputer VisionQUBOQuantitative FinanceMathematicsTransformersMonte Carlo Methods

Timeline

The path so far

  1. May 2026

    Duke Ecology ML Research

    ML/AI Researcher

    Building CNN and transformer-based computer vision models to estimate tree crown cover from PlanetScope satellite imagery, with preprocessing and training pipelines for large-scale geospatial data.

  2. May 2026

    IMC Launchpad

    Quantitative Trading Track — Chicago, IL

    One of 60 students selected for IMC Trading’s Launchpad program. Built and executed trading strategies in simulated markets using probability, expected value, and risk management.

  3. Feb 2026Highlight

    MCM Outstanding Winner

    COMAP Scholar · MAA Award · Leonhard Euler Award

    One of 30 Outstanding Winner teams from 93,977 participants worldwide — then selected as one of 6 COMAP Scholars globally. Paper under consideration for publication in the UMAP Journal.

  4. Aug 2025

    Duke Cosmology Research

    Research Assistant — Scolnic Lab

    Developing Python tools to visualize ~40,000 satellites and debris objects, Monte Carlo collision simulations for low Earth orbit, and a QUBO formulation of satellite routing.

  5. Summer 2025

    EFG Hermes

    Finance Internship

    Early exposure to financial markets and investment analysis at the leading investment bank in the MENA region.

  6. Aug 2024

    Duke University

    B.S.E. Electrical & Computer Engineering + B.S. Mathematics

    GPA 3.966/4.000. Karl Menger Award for outstanding performance in mathematical competitions. Captain of Duke Rugby, BlackRock Asset Allocation Competition finalist.

Projects

Selected work

Simulation, optimization, and machine learning applied to physical and financial systems. Each project has a full write-up.

Research

Satellite Collision Modeling

Monte Carlo simulation of conjunction risk for 40,000+ tracked objects in low Earth orbit.

Monte CarloPythonOrbital MechanicsSGP4

Optimization

Quantum Satellite Routing

A QUBO formulation of satellite routing as a Traveling Salesman Problem, solved with quantum-inspired optimization.

QUBOQuantum OptimizationTSPPython

Machine Learning

Tree Crown Detection

CNN and transformer models for estimating tree crown cover from PlanetScope satellite imagery.

CNNsTransformersComputer VisionRemote Sensing

Mathematical Modeling

MCM Strategic Football Analytics

Outstanding Winner model linking squad decisions to relegation risk and revenue for Crystal Palace FC.

Logistic RegressionPoisson ModelsRevenue OptimizationMCM/ICM

More projects

Smaller builds and experiments, linked straight to the source.

Fake News Detector

NLP classification system for detecting misinformation: Doc2Vec and Word2Vec embeddings plus a BERT-based classifier, with a full data cleaning and preprocessing pipeline.

PythonBERTNLPscikit-learn

Weather Decisions

Probabilistic weather forecasting (HDD/CDD, precipitation, wind) mapped to thin and low-volume market instruments to generate trade signals for proxy baskets.

PythonProbabilistic ModelingMarkets

Market Journal

Daily closing prices for major indices and stocks, paired with curated blurbs explaining the catalysts behind each move. Python backend with automated daily updates.

PythonTypeScriptVercel

Orbit Render

Interactive web-based visualization of orbital mechanics and satellite trajectories — no desktop software required.

JavaScriptVisualizationVercel

Contact

Get in touch

Email is the best way to reach me. Always happy to talk about research, projects, or opportunities.

zain.radwan@duke.eduGitHubLinkedIn