Graduated Dec 2018 University of Pennsylvania, Philadelphia, PA
Master of Urban Spatial Analytics
Master of City Planning
Graduated Jul 2016 Peking University, China
Bachelor of Engineering in Urban Planning
Bachelor of Psychology (Minor)
Sep 2018 - Dec 2018 Data Analyst
Dorothy LLC, Philadelphia, PA
* Created a model to predict flood inundation using Python, providing guidance for homeowners in insurance purchase
* Detected 72% flooded area in 6 counties hit by Hurricane Harvey, 34% higher than FEMA flood maps in the recall by stacking Random Forest, Gradient Boosting and KNN algorithms
* Collected, cleaned, and organized data for 700,000+ grid cells and created spatial features using arcpy module
* Created a demo using Mapbox GL JS to present prediction results to future partners and investors
May 2018 - Aug 2018 Geospatial Data Analyst
China Sustainable Transportation Center, Beijing, China
- Traffic Volume Prediction on National Scale
* Developed a predictive model to estimate the number of lanes by training 100,000+ street segments in 8 cities in China, which served as an important parameter for traffic volume prediction model
* Extracted information from street view using COCO-Stuff and encoded spatial patterns into features using the arcpy module
* Engineered 100+ features and boosted the accuracy in the number of lanes to 70%+ with Python and scikit-learn
* Created Python scripts to automate Geospatial Transformation and Load (GTL) processes, increasing efficiency by 75%
- Empty Running Rate Estimation for Freight Research
* Designed algorithms to estimate the % of deadhead miles using 1,000,000+ location points from a freight-matching app
* Conducted cluster analysis to analyze 1000+ truck drivers’ behavior patterns using K-means and DBSCAN in Python
* Applied Random Forest and Gradient Boosting to predict origins and destinations of freight delivery, with recall rate 80%+
* Identified the deadhead miles using One Class SVM, with the accuracy 70%+
* Built the pipeline for data processing and modeling, and prepared technical tutorials
Jan 2018 - May 2018 Student Researcher in MUSA Practicum
University of Pennsylvania, School of Design, Philadelphia, PA
* Quantified bike share demand for Philadelphia citywide in collaboration with the Bicycle Coalition of Greater Philadelphia
* Performed exploratory analysis and visualized spatial data in R using multiple packages (dplyr, tidyverse, ggplot2, ggmap, etc.)
* Engineered 150+ features using ArcGIS and R, and selected features using Boruta and Random Forest
* Boosted accuracy of the model by 31% using machine learning algorithms, including the OLS, Ridge, Lasso, and Elastic Net Regression, Random Forest, and Gradient Boosting
* Devised an interactive web app to visualize cost-benefit analysis using JavaScript (CSS, HTML, Bootstrap, jQuery)
Sep 2017 - Dec 2017 GIS and Data Analyst
Philadelphia Parks Alliance
* Investigated educational attainment of neighborhoods with census data for 218 recreation centers using ArcGIS
* Designed conservation posters by mapping Philadelphia’s poverty and education conditions
* Mapped financial compliance and structural compliance of recreation centers in Philadelphia using ArcGIS
May 2017 - Aug 2017 Community Outreach Assistant
Philadelphia Parks Alliance
* Collected demographic data of recreation centers and investigated their development and the existing conditions
* Provided support for graphic design using InDesign and Photoshop for community meetings and activities
* Mobilized over 300 community members to attend events through phone-banking and navigated their work
Jul 2014 - Aug 2014 Assistant Planner
BES Tourism Cultural Group
* Devised a tourism product and designed routes of the summer camp for a $150 million Eco-town project
* Researched the potential and market conditions of coastal tourism, as well as designed home port and cruises
* Investigated the market conditions and target groups of senior tourism and outdoor tourism for Yunnan Province
Honorable Mention in ULI Hines Student Competition, 2018
* Achieved honorable mention for our design “Don Ramble” from a pool of 122 participating teams
* Researched local demographic profile and zoning regulations, mapped existing conditions, as well as coordinated design process and financial process
Data Analytics
R(tidyverse, dplyr, ggplot2), Python(pandas, NumPy, SciPy, Plotly, matplotlib), Tableau
Programming
Python, JavaScript (jQuery, bootstrap, Mapbox GL JS), Java
Spatial Analysis
ArcGIS (including ModelBuilder, arcpy module), QGIS, GeoDa, Google Earth Engine
Graphic Design
Adobe Creative Suite (Photoshop, Illustrator, InDesign), AutoCAD, Rhino, SketchUp
Others
NetLogo (agent-based modeling), Arduino (electronic prototyping)