A University of West Florida team placed fifth against teams including Yale University, Dartmouth College and others at a national competition held by the University of South Carolina. The USC National Big Data Health Science Student Case Competition was held Jan. 26-28. UWF graduate students Gail Han and Jason Gerstenberger and UWF undergraduate student Dylan Wright worked for more than 24 hours to solve a healthcare data science challenge. The team, “Data Detectives,” was guided by faculty members Dr. Achraf Cohen and Dr. Shusen Pu, who work in UWF’s Department of Mathematics and Statistics.
“It is essential to recognize that the success of the Data Detectives team not only boosts morale among its members and those in the Computational Statistics and Data Analytics (CSDA) Lab, but also reflects the high quality of education and training provided by our distinguished faculty in UWF’s Mathematics and Statistics Department,” Cohen said.
Students were charged to develop a reliable machine learning classifier to distinguish the discrete segments of EEG data into vigilance states — either paradoxical sleep, slow-wave sleep or wakefulness. EEG data from a small sample of rodents, each having undergone two 24-hour recordings, were provided. Developing mechanisms for automatically and reliably performing the task of vigilance state classification based on EEG would be invaluable to science and, eventually, public health. Students were required to deliver an oral presentation and a written executive summary of their work.
For three months before the competition, the students enhanced their skills by working on data science projects at UWF’s Computational Statistics and Data Analytics Lab. They presented their progress in weekly lab meetings. The lab, known for its focus on machine learning, deep learning and statistical modeling, provided a robust training ground. The team members provided a unique blend of expertise. Han, a first-year mathematical science graduate student focusing on statistics/data science, brought in-depth, hands-on research experience in sleep and neural data analysis. Gerstenberger, a second-year graduate student in mathematical sciences, contributed his expertise in neural network models and data analysis, which he honed through numerous projects. Wright, a senior undergraduate student in mathematics, has studied mathematical modeling of biological systems, which provided valuable insights into data processing and modeling.
“As an undergraduate, I first had a great deal of anxiety about participating in this competition,” Wright said. “Nevertheless, after months of preparation with the CSDA lab and, eventually with my respective team, I acquired the confidence in my ability necessary to accept discomfort when presented with actual data and a problem to solve. I might not have engaged in this competition had it not been for the encouragement and support of Dr. Pu and Dr. Cohen, who promoted student engagement beyond the framework of the conventional classroom environment. I express my gratitude for the opportunity to have teamed up with Jason and Gail.”
A panel of industry and academic experts judged 30 teams’ presentations based on their use of the full analytics process, from framing the problem to methodology selection, data use, model building, and innovation. The University of South Carolina earned first place, Oklahoma State University placed second, College of Charleston was awarded third place, and Dartmouth College received fourth place.
“This achievement highlights our strong commitment to excellence in mentoring both undergraduate and graduate students, as well as conducting research at UWF,” Pu said.
For more information about UWF’s Hal Marcus College of Science and Engineering, visit uwf.edu/hmcse.