DPU DS Team Takes Third at Saint Mary's Data Science Hackathon

The DePauw Data Science Group sent four teams and fourteen people to compete in the Saint Mary's Data Science Hackathon last weekend. The top performing team -- DePauwOutOfBound -- placed third and was awarded $500 for building one of the most accurate predictive models. Other competitors included graduate students from the University of Illinois and undergraduates from Wabash College.

Student utilized a variety of technologies and Data Science methods to build their models. Many took advantage of Python's Natural Language Toolkit to conduct Natural Language Processing on comment data. Algorithms that were implemented included lightGBM (gradient boosting model), random forest regression, and support vector machine classifiers. The event was a first both for Saint Mary's and DPU DS, who had not yet hosted or attended a Hackathon, respectively.