Machine Learning Research Gets Published
The Advances in Science, Technology, and Engineering Systems Journal (ASTESJ) recently accepted an article written by Champlain College faculty members and several current students. The article, “Multi Attribute Stratified Sampling: An Automated Framework for Privacy-Preserving Healthcare Data Publishing with Multiple Sensitive Attributes” will be a part of the 2026 Special Issue on Multidisciplinary Frontiers in Engineering, Computing and Applied Sciences.
Applied Machine Learning and Real-World Impacts
The publication proposes a new framework —“Multi Attribute Stratified Sampling (MASS)” — that tackles the privacy-utility tradeoff in healthcare data publishing through applied machine learning. According to Computer Science & Innovation Assistant Professor Vikas Thammanna Gowda, and the article’s lead author: traditional privacy-preserving techniques require experts to manually specify privacy parameters, often leading to suboptimal results. MASS automates this process by applying machine learning, using classification recall alongside privacy loss and information loss, to evaluate and select the optimal anonymized dataset, ensuring published healthcare data is both strongly privacy-protected and practically useful for downstream predictive modeling tasks.
YinBo Chen ‘27, one of the student researchers on the project, explained that it was easy to isolate and recognize individuals in sets of data — especially if they had a rarer illness. This means there was a gap in both privacy and data collection. The data needed to exist in order to correlate trends, but it needed to protect the individual privacies of the patients that the data is collected from. This team found a way to preserve data collection and analysis without sacrificing usable data.
What started as a classroom exercise became a full research publication, giving all four students hands-on experience taking a problem from concept to a practical, deployable solution — bridging machine learning, data privacy, and software design in a way that goes well beyond what any single course could offer.
Faculty Led Research, Student Driven Progress
Gowda, with funding from a Student Experiences grant, worked alongside four students on the project. Data analytics and data science students Landis Humphrey and Aiden Kadoch, focused on the framework’s development and evaluation over the course of three semesters. In late Summer 2025, YinBo Chen, Computer Science & Innovation, and Olivia Roberts, Data Science, helped get the project over the finish line.
“This research grew organically out of DAT-410 Machine Learning, where students built a framework to classify anonymized healthcare datasets across different privacy settings and observed the classic privacy-utility tradeoff firsthand. Landis and Aiden, eager to go beyond the coursework, wanted to tackle the algorithmic side of data privacy and apply their Computer Science minor, which led to the design of the MASS algorithm itself. Once the framework was complete, YinBo and Olivia joined the project, contributing object-oriented data visualization modules that brought the results to life.“ Gowda stated.
Chen, a Computer Science and Innovation Major, was invited to join the project after he took a data analytics class with Vikas and showed interest. According to him, the ultimate goal was to find a solution on making healthcare more private. Chen’s role focused on coding and the visualizations of the datasets. These graphs, made through python coding, were extremely valuable to help see the results of their work.
“I already knew some of the basic concepts [from Vikas’ class],” Chen said, “This project only helped me expand [that knowledge] further.”
Each week, the students received an assignment that added onto the work from the previous weeks on the project. This created an iterative process, where they were constantly building upon the project. “It was rewarding to actually see the results in front of me.” YinBo stated.
Ready to Dive Deeper into Applied Machine Learning?
Check out some of the programs at Champlain where you can start building your own software from day one.
Author
More Inside The View
Ideas
From the minds of our students, faculty, and alumni.
News
The latest from Champlain College.
People
Champlain is more than just a place; it's a community.
Places
On campus, in Burlington, and beyond.
Events
Check out our many campus events and get involved! Refine your search by using the filters or monthly view options.