Behaviour Algorithms Prediction
This project marked my first step turning raw visual observation into measurable quantitative behavioural evidence. Two-photon equipment records a mouse in an open-field arena; I built the pipeline that turns the video into trajectory data, annotated ear-deflection features, and ran the analysis in Python.
The output was not a pretty chart but a behaviour model: how the animal moved between regions, how its head orientation shifted, and how those features separated experimental groups. Everything downstream in this portfolio — the statistical rigour, the habit of showing my working — starts here.
Perceived Social Mobility & Humility
This is where I shifted focus from animal behaviour to human judgement. Conducted under Peking University's Yuanpei College, the study designed online experiments exploring the relationship between social mobility and humility.
I collected 193 valid responses, then wrote and defended a 2,005-word paper examining how social context and reference groups shape individual decisions and perceived rationality. The research question asked not what people believe, but what makes a belief feel reasonable to them.
Medical Overuse & Web Game
Working with two teammates, we combined four different subjects and wrote a complete research paper. I learned basic web coding on my own to make the game, then presented our findings.
This project marks my shift from academic writing toward public-oriented outreach. By analysing the root causes of medical overuse, I built an interactive web-based game that raises public awareness of unnecessary medical treatment.
The live build below is the real thing — not a mock-up. Try it: every consultation you take is a decision the game is asking you to defend.
Click to enlarge
Medical Overuse · research poster · original size
Also live at mistymedicalmaze.netlify.app
U.S. Stock Market Analysis
I built a regime-adaptive investment framework and backtested ETF performance across historical market cycles — then produced a documentary to visualise the results rather than leave them in a spreadsheet.
This is where I expanded my inquiry from individual choices to collective crowd behaviour. The framework asks a simple question: when the regime of the market changes, does the strategy still hold?
Click to enlarge
Analysis poster · full resolution
Consumer Preference Inquiry
I handed out questionnaires alongside introduction cards with distinct thematic framings to people during my UPenn summer programme, and collected 130 responses. Then I analysed the feedback and visualised the results to compare which narrative resonated most strongly with potential buyers.
This project shifted my inquiry from market behaviour to how targeted narratives shape individual consumer choices. The same product, told four ways.
Each framing was printed on a card and handed out with the questionnaire. The three illustrated boards below are the framings I drew myself — the same batik, the same product, argued three different ways.
AP Seminar IWA / IRR
This project marks where I consolidated all prior inquiry habits into formal academic argumentation. I completed an individual written argument and an individual research report exploring layered perspectives on a complex social issue.
Four documents below: the Individual Written Argument on the impact of commercialised cafés and its two counter-readings, the research report on PM2.5 and adolescent mental well-being, the full draft, and the final submitted paper.
