Work
Projects4
Treehole ↗
An anonymous sharing community with email-code sign-in, posts, comments and moderation tools, backed by a hosted Postgres database.
Learnware Benchmark
At Nanjing University's LAMDA group: a benchmark for finding and reusing pre-trained models. 30 base models across 10 UCI/OpenML datasets, statistical specifications for each, and 4 identification baselines compared on accuracy, runtime and reuse performance.
About
Hi! I'm Eloise (Xitong He), a Statistics student at the University of Washington, expected to graduate in 2029 (GPA 3.66).
I'm a data analyst at Solid Ground, a Seattle nonprofit, where I turn housing service records into insights and dashboards that help place people in homes. This past summer I worked as a research assistant in Prof. Zhi-Hua Zhou's LAMDA group at Nanjing University, building a benchmark for reusing machine-learning models.
I've also spent years teaching — calculus, statistics and TOEFL — and I enjoy explaining ideas clearly. Outside of data, I design clothing inspired by ethnic minority cultures, and one of my series was published in a French fashion magazine.
Experience
Research & Data
Data Analyst · Solid Ground
- Analyzed 1,000+ housing service records in Python and SQL to find service gaps and improve placement strategies.
- Supported housing placement for 200+ people by analyzing referral patterns, client needs and service access.
- Cut reporting time by 30% with automated data cleaning and reporting workflows.
- Built Tableau dashboards tracking housing outcomes and referral success rates.
Research Assistant · LAMDA Group, Nanjing University
- Built an experimental benchmark for learnware identification and reuse across 10 public tabular datasets.
- Trained 30 base models (Random Forest, SVM, XGBoost) and generated statistical specifications of each model's data distribution.
- Implemented model-similarity matching and compared 4 baselines; wrote the technical report.
Research Assistant & Vice President · Ethnic Culture Research Society
- Analyzed cultural survey data on Xinjiang ethnic groups; organized cross-cultural events that raised student participation by 30%.
Teaching
TOEFL Tutor · Private tutoring
- Tutored high school and college students one-on-one with study plans built around each student's level and target score.
Peer Tutor, Calculus & Statistics · School Academic Support Program
- Selected by faculty to tutor peers one-on-one and in small groups for coursework and exam preparation.
Education & Awards
B.S. Statistics · University of Washington
- GPA 3.66. Coursework: Data Programming, Intro to Data Science, Statistical Methods, Matrix Algebra, Ethics of Algorithms.
First Prize · Global Hi-MCM Mathematical Modeling Competition
- Among 500+ international teams. Also a national finalist in the CTB Mathematical Modeling Competition, using Python and R.
Independent Designer
- Designed a clothing series inspired by ethnic minority cultures; published in the French fashion magazine L'attirance.
Contact
Let's talk
Open to internships, collaborations, or just a chat about something you're building.