Hi, I'm Zijie!
I am a student at the University of Michigan majoring in Data Science. I'm a passionate learner who is always searching for my next obsession.
I love solving complex problems—whether it's optimizing algorithms, simulating pirate treasure hunts, or analyzing wildlife data to uncover hidden patterns. Right now, I’m diving deeper into AI and machine learning, with the goal of building intelligent systems that make a real-world impact.
Some relevant coursework I've taken / am taking are: Discrete Math, Data Structures and Algorithms, Introduction to Probability and Statistics, Practical Data Science, Appplied Regression, Web Systems, and Intro to Machine Learning.
When I’m not staring at my laptop, I’m probably out taking pictures or browsing the latest mechanical keyboards.
Analyzed early-game metrics to identify key predictors of match outcomes, focusing on the first 15 minutes of professional games.
Built a predictive model using a dataset of 10,000+ pro matches, achieving 74% accuracy in classifying match outcomes after optimizing model hyperparameters.
Utilized Python, Pandas, NumPy, and scikit-learn for data preprocessing, feature engineering, and model development.
Made a minimalistic personal website from scratch using HTML and CSS. Includes interactive features and mobile support.
Used Python, pandas, NumPy, Matplotlib, seaborn and sklearn to analyze and visualize data of the wolf-moose population on Isle Royale.
Performed regression modeling, confidence interval tests, hypothesis tests, and exploratory data analysis to predict certain parameter trends with 65-67% accuracy.
Programmed Arm, Drivetrain, Flipper, Auton and ShuffleboardManager subsystems and commands for our high school FRC team's robot.
Used Java and WPILib.
Research Assistant
East Lansing, MI
May 2025 – August 2025
Created object-masked evaluation pipelines by combining segmentation outputs from MaskDINO into unified foreground masks for monocular depth estimation analysis on KITTI and DDAD datasets. Performed range-based evaluations of Metric3D and UniDepth by analyzing performance across different depth intervals using benchmark metrics, discovering a fall off in model performance as distance increases past 20-30 meters. Compared full-image and object-masked evaluations on KITTI and DDAD, revealing trade-offs in accuracy between global scene estimation and foreground objects.
Research Assistant
Ann Arbor, MI
May 2025 – August 2025
Gathered and organized data by compiling distances from cement plants to neighboring states and their largest cities into a structured dataset using a Python script with OpenRouteService API, improving data accuracy for further analysis. Implemented a greedy minimum-cost flow algorithm to model and optimize cement distribution across states based on transportation costs, finding ways of reducing carbon emissions from cement transportation and enhancing environmental sustainability.
Data collector and analyst
East Lansing, MI
Nov 2022 – Oct 2023
Wrote a research paper on the use and effectiveness of honey alternatives for overwintering bees and entered Regeneron Science Talent Search under Prof. Zachary Huang.
Summer High School Academic Program for Engineers (SHAPE)
New York, NY
Jul 2022
Worked on a collaborative project to design, build, and test a fully autonomous robot from scratch that navigates a maze to put out a randomly placed candle under Prof. Sinisa Vukelic.
Brookhaven National Laboratory High School Research Program (HSRP)
Virtual
Aug 2023
Worked under Dr. Enyuan Hu to write a research poster about characterizing battery cathode materials by using X-ray absorption spectroscopy.
Host – Part-time
East Lansing, MI
Apr 2024 – present
Bussed tables, wiped windows, cleaned bathrooms, managed reservations and waitlist. Greeted and seated customers, answered phone calls, and completed all to-go/online orders.
C++
Python
Java
HTML
CSS
pandas
NumPy
sk-learn
jupyter
Github
VS Code