Yuexing Hao's Projects
To provide a tangible showcase of my work, I’ve included a portfolio of my most recent projects. Take a moment to explore below, and reach out if you’re interested in learning more about the projects .
Vita Mask Software
Vital Mask is a diagnostic support system designed for Emergency Departments (EDs) that monitors respiratory rate, heart rate, blood oxygen (SpO2) levels, and body temperature in real-time. Vital Mask software innovatively implements different scoring systems: NEWS, NEWS2, and MEWS to resolve the ED's overcrowded problems. Company Website
Clinical Decision Support Tools (DSTs)
Pallia Care
Pallia Care is an application specifically designed for palliative patients and their families. It visualizes each treatment option and helps patients and their families to make efficient and optimized decisions on treatment plans. Pallia Care successfully bridges the gap between physicians and patients. demo
OOTOMS
OOTOMS is an outcome-oriented techniques for user’s optimized model selections using visualization tools. We built two generalizable visualization techniques for model selection under the outcome-oriented framework: outcome-first technique and stage-first technique. These two techniques reverse the traditional overview-first technique and provide users visual aids to choose the “best” option given several models.
Agriverse
Agriverse is an abbreviation for agriculture metaverse. Our project aimed to save water on dairy farms by building an intellectual sprinkler system, employing machine learning / Image Processing techniques. Our team won the "best use of data"($1500) in Cornell Digital Agriculture Hackathon. I also presented this project at KDD 2022 AIoT Workshop, sponsored by Microsoft. Link to our pitch video.
Gin Rummy AI Player
Gin Rummy Card game is an imperfect information card game that allows players to create strategies and mathematical calculations to maximize their win rate. In this paper, we presented an AI Gin Rummy player by using Java. Based on the game’s rules, we developed both a discard algorithm and a draw algorithm to maximize the win rate by switching different strategies. Moreover, to win more points, we developed dynamic strategies, which can be more responsive and intelligent. The strategies’ algorithms are dynamic based on previous games’ results. By implementing discard and draw algorithms, the win rate increases from 50% to 57.85%. Including our dynamic strategies further increase the win rate to 67.735% (among 100,000 games).
POSER
For this project, I designed an effective sitting posture modification application designed for sedentary users. Poser employed webcams to keep track of users’ postures by using computer vision technology and sending alerts to adjust users’ sitting postures.
Story Illustration
I assisted in research support of data processing operations for large survey research projects. In this project, I displayed visual story illustrations from the consequences of texts, by using machine learning languages in Python like PyTorch. link
ICM Meritorious Prize
I worked individually for the interdisciplinary contest in modeling (ICM) in 2020 and won the meritorious prize. link