Mines Data Science Team Continues Success Streak at Regional Competition in Minneapolis

A team of South Dakota Mines mathematics and computer science students earned a top honor at the recent MinneMUDAC data science competition, extending their winning streak at the annual event that challenges graduate and undergraduate teams to tackle real-world data and demonstrate their skills.

Each year, the competition offers a new challenge. This year, event organizers MinneAnalytics teamed up Big Brothers and Big Sisters Minnesota, the largest and oldest youth mentoring organization in the greater Twin Cities, to focus on early interventions to promote long-lasting and successful pairings between Bigs, caring adults who volunteer as mentors, and Littles, kids between 8 and 13 and their families.

The Mines undergraduate team of math and computer science majors Angelo Cosato, Liz Herting, Eric Johnson, and Tammy Nguyen took the prize for best prediction out of all 35 undergraduate teams in the competition. “They were able to predict the mentor match length within seven months,” said Kyle Caudle, Ph.D., a professor of mathematics at Mines and one of the team’s advisors. “It is excellent when you think that matches typically last anywhere between one month and 90 months.”

The team achieved a Root Mean Square Error (RMSE) of 7.7, closely matching the most accurate graduate model with an RMSE of 7.5. This shows that the Mines undergraduate math and computer science students performed on par with graduate students from other universities, added Travis Kowalski, Ph.D., head of the Department of Mathematics.

“This also means that the Mines Data Miner teams have never returned from a MinneMUDAC competition without winning at least one of the modeling or presentations prizes,” Kowalski said.

Along with the Data Miner team’s success at the annual competition, the university has twice been selected for a $1,000 student scholarship in recognition of its engagement with the data science community and its dedication to its students in the data sciences.

Everett Sandbo, a junior math and computer science major, was awarded this year’s $1,000 scholarship. Nominated by Caudle, Sandbo played a key role in updating TensorTools, an R software package that offers foundational tools for tensor operations.

“With a strong interdisciplinary focus, Everett is passionate about tackling machine learning challenges and has actively contributed to faculty research projects,” Kowalski said.

Sandbo is also on the Mines Hardrocker cross-country and track and field teams.