• 🎓 Sophomore at Grinnell College, pursuing B.A. in Computer Science & Mathematics (Concentration: Statistics)
• 📊 Striving for jobs in quantitative research, data engineering, and AI/ML
• 💻 I was a Data Engineer and AI Engineer at Gtel Data Research Group in Summer 2025, and a NLP Intern at Data Glacier in Fall 2025.
• 🌱 Learning diffusion models, reinforcement learning, and LLM finetuning
• 👯 Open to collaborating on quant research, machine learning/computer vision projects, and solving sudoku problems
• 💬 Ask me about machine learning, deep learning architectures, or just life in general
• 📝 Portfolio & Blog: https://ducduong-portfolio.vercel.app/
• ⚡ Fun fact: I love dabbling in variants Sudoku, badminton, and soccer

• LinkedIn
• School Email: [email protected]
• Work Email: [email protected]

• 🧩 Cracking the Cryptic is the best YouTube channel in the world.
• ⚽ Born to play soccer but peaked at 🏸 badminton
• 📖 Reading AI/ML research papers, quant finance literature, and manga

🔗 Sponsored by American Statistical Association (ASA) & CAUSE. | Dec. 2025

Analyzed theft patterns across Los Angeles using the 2020 LAPD dataset to understand how spatial and demographic factors affect theft distribution.

• Applied nested logistic regression models with predictors such as population size, density, victim age, sex, and race.
• Found population density to be the strongest negative predictor of theft, while demographic analysis showed older victims and women were slightly more likely to be targeted.
• Highlighted racial differences in exposure to theft vs. violent crimes.
• Work was recognized nationally, earning 1st Prize in the USPROC Introductory Statistics Class Project competition.

🔗 GitHub Repository | Aug 2024 – May 2025

A research project exploring whether machine learning models can distinguish tonal vs. non-tonal languages from multilingual audio samples.

• Collected and processed 125 multilingual audio clips from 18 countries.
• Designed spectral and pitch-based features that reduced raw noise by 30% and improved dataset balance.
• Benchmarked 7 ML models (logistic regression, SVM, random forest, neural nets, etc.) with cross-validation, achieving 65% accuracy (20% over baseline).
• Built reproducible pipelines in scikit-learn and PyTorch for comparative metrics (precision, recall, F1).
• Proposed scalable data collection strategies for future interdisciplinary research in linguistics + machine learning.

• 🏆 1st Prize – USPROC Introductory Statistics Class Project Competition (June 2025)
  Project: Spatial And Demographic Effects On Theft Distribution Across Los Angeles
• 🏆 2nd Place – 2025 Iowa Collegiate Mathematics Competition (99/100 score)
• 📜 Machine Learning and Data Science A–Z (Python/R, Udemy)
• 📜 UR2PhD Undergraduate Pre-Research Experience Course Credential