This fall, we invited undergraduate and graduate students to explore real-world healthcare data for MinneMUDAC 2017. Last Saturday, Nov. 4, more than 60 teams gathered in Eden Prairie to present their results to judges from the analytics community and compete for cash prizes.
This year teams were given access to an HIPAA-compliant, de-identified, limited dataset including health insurance claims data for individuals 18-45 years of age with claims evidence of Type 2 diabetes. The challenge question required students to describe the population of persons with claims evidence of diabetes and compare these findings among important subgroups in the population.
Advanced teams were also required to create a predictive model identifying a group of persons with claims evidence of Type 2 diabetes and health care experiences that have an increased probability of becoming a high-cost Type 2 diabetes patient in the next 12 months.
Judges evaluated teams based on completeness of answering the challenge questions, as well as data preparation, communication of ideas, creativity and appropriateness of methods. In addition, advanced teams were evaluated on the accuracy of their predictive models.
A team from the Carlson School of Management received First Prize Overall in the Advanced Graduate Division. Judges were impressed by the background and insights provided by the team, as well as the interdisciplinary mix of team members. The team included a health policy management doctoral student.
“We are seeing blended teams succeeding in business, and many student teams are creating the same and winning,” said Dan Atkins, co-founder and executive director of MinneAnalytics. “Last year an English major helped propel her team past the competition to win. Having the best answer doesn’t do it, you need to understand the domain and best communicate an answer.”
Advanced: Graduate Division
For teams with at least one graduate student with advanced data management, data programming, and statistical/analytic skills to support predictive modeling.
1st Prize University of Minnesota Carlson School of Management – Team Best Idea
Acumen Award South Dakota State University – Team SDSU Grad Gold
Discovery Award University of Minnesota – Team Positively Skewed
Honorable Mention University of Minnesota – Team Gopher Analysts
Honorable Mention University of Minnesota – Team Sunshine, Lollipops and Rainbows
Advanced: Undergraduate Division
For undergraduate teams with advanced data management, data programming, and statistical/analytic skills to support predictive modeling.
1st Prize Winona State University – WSU Team 2
Acumen Award Winona State University – Team Deppa’s Kids
Discovery Award University of Minnesota Duluth – Team We R SASsy Dogs
Honorable Mention Winona State University – WSU Team 1
Honorable Mention Winona State University – Team Paulson’s Pals
Honorable Mention Macalester College – Team The Shu-Men
For students early in their studies who have limited experience and have novice to intermediate data management, data programming and statistical/analytic skills.
1st Prize Gustavus Adolphus College – Team Golden Data Analysts
Acumen Award Minnesota State University, Mankato – Team MavAnalytics
Discovery Award Metropolitan State University – Team Alpha United
Thank you to all of the participating students for your hard work and dedication, as well as the faculty and staff advisors who provided support. We would also like to thank the more than 160 professionals who served as judges and shared their experience with the next generation of data scientists, and the volunteers who made this competition happen.
MinneMUDAC is an annual data science challenge presented by MinneAnalytics in partnership with the Midwest Undergraduate Data Analytics Competition. Thank you to MinneAnalytics members and sponsors for making these events possible.