Four students from the Master of Science in Analytics (MSA) program took first place at the 2021 Humana-Mays Healthcare Analytics Case Competition, a partnership between Mays Business School at Texas A&M University and the health and well-being company Humana Inc. The team includes M.S. Analytics students Siyan Cai, Manqiu Liu, Tsz Fung Pang, and Jia Shi.

Georgia Tech’s MSA is an interdisciplinary program that leverages the strengths of the H. Milton Stewart School of Industrial and Systems Engineering, the College of Computing, and the Scheller College of Business. The case competition was open to students from master’s programs in business, healthcare, or analytics. More than 750 students representing 75 major universities across the country competed, and the winning team received the first-place prize of $50,000 following a virtual presentation to an executive panel of judges.

The challenge of the case competition was to minimize health inequities and address vaccine hesitance among vulnerable and underserved populations using the power of data analytics. Students were asked to utilize prescriptive and predictive modelling to predict vaccine hesitancy in specific segments, and accordingly propose targeted outreach to remove the barriers of the hesitant populations to receive vaccinations.

“This case focused on a very distinctive segment of the population in the healthcare industry, so it was difficult at the beginning to grasp the problem, especially since none of us had any domain knowledge,” said Cai.

However, the team was not intimidated, and they performed a lot of research on the topic, consulting with field experts and conducting many discussions and brainstorming sessions. After understanding the context and central problem to solve, the students broke the case into three key questions to be solved. These questions laid a foundational framework for their analysis and strategies; as they moved forward, they made sure to always address the questions at the heart of their solution.

One of the challenges the team faced was dealing with a very large dataset, which included approximately one million records of Humana Medicare members with more than 400 features. During the data preprocessing, they extensively leveraged data manipulation techniques, ending up with a model that ranked top five in the leaderboard.

The main reason why the team’s model has high predictive power is that they performed well-rounded data preparation. Rather than merely applying modeling techniques to fix the problem, the students focused on the fundamental causes of bias in the dataset as well as alternative solutions.

“We wanted to aid the disadvantaged and find solutions to address vaccine disparity and inequity,” said Liu. “We believe that data science has the potential to benefit society, and this competition is an excellent opportunity for us to see what kind of beneficial influence analytics may have.”