[Machine Learning | GUI] Created an open-source Python-based computer vision toolkit (UrbanTrace) to detect urban changes at the building level using high-resolution aerial imagery, incorporating deep learning models (YOLO, Siamese Networks), and integrated the tool into a QGIS plugin with PyQt.
[Computer Vision] Processed and analyzed large-scale geospatial datasets (6.5M buildings across 35,000+ square miles) using Python, PyTorch, OpenCV, and GIS frameworks.
[LLM] Use Large Language Models via LangChain—including GPT-4, Claude 3, Qwen2, and LLaMA 3—to conduct sentiment analysis on 10,000+ Twitter/X posts, evaluating public perception of New York City’s congestion pricing plan.
[Causal Inference] Use logistic regression, spatial econometric analysis, and an explainable machine learning model (XGBoost+SHAP) to examine urban development trends and demographic shifts with various Python packages
[Supervision] Provided data analysis (Python and R), machine learning, and GIS support to lab researchers and supervised undergraduate and master’s students.
Transportation Planner
HDR, Inc.
AI and Data Related:
[ETL] Developed a Python + VBA pipeline to warehouse, analyze, and visualize transportation data from various sources (MySQL, CSV, OData) across the NY/NJ region
[MySQL] Designed and developed a Windows tablet survey application using Python and PyQt, enabling surveyors to collect 5000+ questionnaires with automated daily uploads to a central MySQL server.
[Data Visualization] Built interactive live data dashboards in Tableau, integrating MySQL databases to visualize transportation analysis results.
[Workflow Automation] Automated data analysis and visualization by developing VBA macros in Excel and Adobe Illustrator.
[Data Collection] Supervised 10+ third-party traffic data collectors in the field and ensured data quality
Planning and Design Related:
Develop traffic models to assess the environmental impacts of various types of developments.
Supervised 10+ third-party traffic data collectors in the field for data collection.
Education
PhD in City Planning
University of North Carolina at Chapel Hill
Thesis: “Explore Urban Regeneration Pattern, Drivers, and Social Impacts in the Southeastern United States Through AI-Powered Analysis” . Supervised by Prof Yan Song. Presented papers at planning journals and AI conferences.
University of North Carolina at Chapel Hill ∙
October 2022
Social Science Track: “Flood? Let’s Help! San Francisco Flood Health Vulnerability Analysis”
Analyzed the influencers of flooding risk in San Francisico with KNN, Random Forest and SHAP.
Transportation Leadership Fellows
University of North Carolina at Chapel Hil ∙
January 2020
China National Undergraduate School Social Study Competition