Technologies: Python, Jupyter Notebooks
Skills & Methodologies: Unsupervised Learning (Clustering)
GitHub Repository
Professional sports leagues in the United States such as the MLB, NFL, and NBA require teams to undergo large amounts of travel. Stakeholders would like this travel to be minimzed as doing so reduces costs and environmental impact while improving player recovery. Teams most often play opponents within the same division. Thus, if teams within the same division as are close as possible overall travel across the league can be minimzed. One way of solving this problem is through unsupervised clustering techniques such as K-Means. However, divisons must be approximatelty the same size. Thus, I employ a modified K-Means algorithm to create divisions which minimze travel while retaining uniform size. I published an article in Towards Data Science detailing my approach and results. A link to this article can be found in the GitHub repository or under the Writing Samples tab on this website.