Urban Spatial Analytics  ·  Data-Driven Planning 城市空间分析  ·  数据驱动规划

Xinyuan
Christine
Cui

I use spatial data to understand how cities work — from how people move, to how neighborhoods are valued, to how we build more equitable urban environments. 我用空间数据理解城市的运作方式——从人群流动,到街区价值,再到如何构建更公平的城市环境。

Christine Cui
Philadelphia · 2026 费城 · 2026
4.0/4
GPA at Penn 宾大绩点
9+
Projects 项目
3×
Natl. Awards 国家级奖项
01 · About 01 · 关于我

Cities are puzzles.
Data helps us solve them.
城市是谜题,
数据帮我们解开它。

Hello — I'm Christine, an urban planner and spatial analyst who believes the best city interventions start with rigorous, honest data. I'm a graduate student in Urban Spatial Analytics at the University of PennsylvaniaGPA 4.0, working at the intersection of machine learning, GIS, and urban policy.

My work spans road safety prediction in Philadelphia, housing tax equity modeling, benchmarking generative AI for satellite imagery, and building cloud-based tools that help planners see what they couldn't see before. I approach every project with equal attention to technical precision and human impact.

Before Penn, I studied Urban and Rural Planning at Tianjin University and spent a semester at The University of Hong Kong. I've interned at two of China's leading planning institutes — Shenzhen Urban Planning & Design Institute and Shanghai Tongji Urban Planning & Design Institute.

Actively seeking full-time roles in urban data analytics and planning technology — available from May 2026.

LinkedIn ↗ ·  GitHub ↗ ·  cui12@upenn.edu

你好,我是崔馨元(Christine),一名城市规划师与空间分析师,相信最好的城市干预措施始于严谨诚实的数据。我目前就读于宾夕法尼亚大学城市空间分析专业GPA 4.0,在机器学习、GIS与城市政策的交叉领域开展研究。

我的研究涵盖费城道路安全预测、住房税收公平建模、生成式AI卫星影像云去除,以及构建帮助规划师洞察城市问题的云端工具。每个项目中,我都以同等的技术精准度和人文关怀参与其中。

加入宾大之前,我在天津大学学习城乡规划,并在香港大学完成了一学期的交换学习。曾在国内两家顶尖规划机构实习——深圳市城市规划设计研究院和上海同济城市规划设计研究院。

目前正积极寻求城市数据分析与规划科技方向的全职职位,2026年5月起可入职。

LinkedIn ↗ ·  GitHub ↗ ·  cui12@upenn.edu

Education 教育背景
2025 – 2026
University of Pennsylvania
Master of Urban Spatial Analytics · Philadelphia 城市空间分析理学硕士 · 费城
GPA 4.0 / 4.0
Jan – May 2024
The University of Hong Kong
Exchange Program · Urban Studies 交换生项目 · 城市研究
2019 – 2025
Tianjin University
Bachelor of Engineering · Urban and Rural Planning 工学学士 · 城乡规划
Experience 工作经历
Sep – Oct 2024
Shenzhen Urban Planning & Design Institute
Urban Design Intern · Housing planning, policy analysis, urban renewal finance 城市设计实习生 · 住房规划、政策分析、城市更新融资
Jul – Aug 2023
Shanghai Tongji Urban Planning & Design Institute
Planning Intern · Land use analysis, economic metrics, bio-industry zoning 规划实习生 · 土地利用分析、经济指标测算、生物产业用地规划
02 · Projects 02 · 项目

Selected Work 精选作品

Hover to preview · Click to explore each project. 悬停预览 · 点击查看项目详情。

01
Capstone 毕业设计 Ongoing 进行中
Off-Peak Roadway Safety 非高峰期道路安全
Understanding Off-Peak Roadway Safety 理解非高峰期道路安全
Predictive model + spatial analytics tool identifying high-speed driving risks in Philadelphia. Dashboard to help city planners mitigate off-peak fatalities by modeling road design relationships. 识别费城高速驾驶风险的预测模型与空间分析工具,通过建模道路设计关系,为城市规划师提供非高峰期事故防控辅助决策支持。
R Spatial Modeling Dashboard Road Safety
View Project ↗ 查看项目 ↗
Jan 2026 — Ongoing Capstone Research 毕业论文研究
02
Cloud · ML Ongoing 进行中
Philadelphia CAMA System 费城房产税评估系统
Philadelphia CAMA Property Tax 费城房产计算机辅助批量评估
GCP pipeline (Cloud Storage, BigQuery, Cloud Functions, Cloud Run) for automated property valuation. ML models predict assessed values, deployed via interactive map dashboard. 基于GCP管道(Cloud Storage、BigQuery、Cloud Functions、Cloud Run)的自动化房产估值系统,ML模型预测评估价值并通过交互式地图仪表板展示。
BigQuery Cloud Run Python Mapbox
In Progress ↗ 进行中 ↗
Mar 2026 — Ongoing Data Engineering 数据工程
03
AI / ML Complete 已完成
Generative AI for Cloud Removal 生成式AI卫星云去除
Generative AI for Satellite Cloud Removal 生成式AI用于卫星影像云去除
Benchmarking GANs & Diffusion models vs. multi-temporal methods on 15,000 Sentinel-2 patches at 5–70% occlusion. GAN achieves 32.76 dB PSNR vs. 15.09 dB temporal baseline at high coverage. 在15,000张Sentinel-2卫星图像(5–70%云覆盖率)上对GAN、扩散模型与多时相合成方法进行基准测试。高覆盖率下GAN达到32.76 dB PSNR,远优于时序基准15.09 dB。
GAN / Diffusion Remote Sensing PyTorch Sentinel-2
View Project ↗ 查看项目 ↗
Mar – Apr 2026 Machine Learning 机器学习
Safe Passage residuals map
04
Spatial Econ 空间计量
Safe Passage: Crime at SEPTA Stops 安全通道:SEPTA站点犯罪分析
Safe Passage: Crime Risk at SEPTA 安全通道:SEPTA站点犯罪风险
Negative Binomial spatial risk model across 5,884 stops. Pseudo-panel fixed effects isolate ridership's causal impact. 56% MAE reduction via 5-fold CV and Moran's I validation. 5,884个SEPTA站点的负二项空间风险模型,伪面板固定效应剥离客流量因果效应,通过5折交叉验证和Moran's I空间验证,MAE降低56%。
R NegBin Model Moran's I Philadelphia
View Project ↗ 查看项目 ↗
Sep – Dec 2025 Spatial Econometrics 空间计量经济学
Housing sales price quartile map
05
Equity 公平性
Path to Parity: Fair Housing Valuation 平等之路:公平住房估值
The Path to Parity 平等之路
Housing valuation in R: ZIP Code fixed effects + KNN spatial features. R² = 0.84 on 31,000+ properties. Residual maps reveal systematic bias in misvalued neighborhoods. R语言住房估值模型:邮编固定效应+KNN空间特征,在31,000+套房产上R²=0.84,残差地图揭示被低估社区的系统性价值偏差。
R KNN Features 10-fold CV Housing Equity
View Project ↗ 查看项目 ↗
Sep – Oct 2025 Predictive Modeling 预测建模
Yunnan Odyssey dashboard
06
Live Site 在线演示
Yunnan Odyssey Travel Platform 云南漫游旅行平台
Yunnan Odyssey: Plan Your Myth 云南漫游:规划你的传奇
Web travel platform bridging emotional inspiration with rational logistics. Integrates social discovery, route optimization & budget tracking via Leaflet, Firebase, and Mapbox. 连接情感灵感与理性规划的网页旅行平台,通过Leaflet、Firebase和Mapbox整合社交发现、路线优化与预算追踪功能。
JavaScript Mapbox Firebase Leaflet
View Live Site ↗ 查看在线演示 ↗
Oct – Dec 2025 Web Development 网页开发
Tianjin metro station classification map
07
Mobility 城市交通
Bike-Sharing & Rail Integration 共享单车与轨道交通接驳
Bike-Sharing Integration with Rail Transit 共享单车接驳轨道交通研究
HDBSCAN clustering defines bike-metro service areas and station typologies. Quantitative modeling of built-environment factors driving first-/last-mile connectivity. HDBSCAN聚类识别单车-地铁服务范围与站点类型,对驱动首末公里接驳的建成环境影响因素进行定量建模研究。
Python HDBSCAN First-Mile GIS
View Project ↗ 查看项目 ↗
Oct – Dec 2024 Urban Mobility 城市交通
Shanghai residential happiness distribution map
08
XGBoost
Community Environment Optimization 社区环境优化研究
Community Environment & Resident Happiness 社区环境与居民幸福感
XGBoost + TreeSHAP interpretability analyzing social, economic, and ecological indicators affecting residential happiness across high- and low-efficiency areas via TOPSIS. XGBoost+TreeSHAP可解释性分析,研究影响居民幸福感的社会、经济与生态指标,通过TOPSIS方法识别高低效能社区并提出差异化优化策略。
Python XGBoost SHAP TOPSIS
View Project ↗ 查看项目 ↗
Apr – Jun 2024 ML Interpretability 机器学习可解释性
Urban emotion spatio-temporal analysis cover
09
Time Series 时序分析
Spatio-temporal Urban Sentiment 城市情感时空分析
Spatio-temporal Urban Sentiment Analysis 城市情感时空分析
Weibo check-in + Baidu Index time series for Beijing & Shanghai. STL decomposition and stationarity testing reveal how urban events register in collective digital expression. 结合北京、上海的微博签到数据与百度指数构建时间序列,通过STL分解与平稳性检验揭示城市事件如何在集体数字情绪表达中留下印记。
Python / R STL Decomp Time Series NLP
View Project ↗ 查看项目 ↗
Oct – Dec 2023 Sentiment Analysis 情感分析
03 · Skills 03 · 技能

Tools of the Trade 专业工具与技能

Programming & Analysis 编程与分析
Languages 编程语言
R — tidyverse, sf, tidycensus
Python — pandas, scikit-learn, PyTorch
SQL · PostgreSQL · PostGIS
JavaScript
Statistical Methods 统计方法
Spatial Econometrics (SLM, SEM, NegBin) 空间计量经济学(SLM, SEM, NegBin)
Machine Learning & Cross-Validation 机器学习与交叉验证
Time Series & STL Decomposition 时间序列与STL分解
Spatial & Design Tools 空间与设计工具
GIS & Cloud GIS与云平台
ArcGIS Pro
GeoDa · Fragstats · SPSS
Google Cloud Platform · BigQuery 谷歌云平台 · BigQuery
Visualization & Design 可视化与设计
Mapbox Leaflet Kepler.gl AutoCAD SketchUp Illustrator Photoshop Enscape Ecotect
Languages 语言能力
Chinese (Native) 中文(母语)
English (Fluent) 英语(流利)
04 · Recognition 04 · 荣誉

Honors & Awards 荣誉与奖项

2020 · 2023
Outstanding Student Leader 优秀学生干部
Tianjin University 天津大学
Jul 2023
Yuanye Awards for Students — Third National Prize 元野奖学生竞赛 · 全国三等奖
National Level Competition 全国性竞赛
Mar 2023
Milan Design Week China Collegiate — Second National Prize 米兰设计周中国高校竞赛 · 全国二等奖
International Competition 国际竞赛
05 · Contact 05 · 联系方式

Let's connect & build 期待与您联系合作

Whether you're hiring, researching, or want to talk cities and data — I'd love to hear from you. 无论您是招聘、研究,还是想聊城市与数据——欢迎随时与我联系。