Fantasy Sports: Unsupervised Learning Project

OnSports is a fictitious Fantasy Sports platform that has fantasy leagues for different sports. For each player, a price is set at the start, and the price keeps changing over time based on the performance of the players in the real world.

In this MIT PE Project, I conducted a cluster analysis to identify players of different potentials based on previous season performance. My objective is to help Onsports understand the patterns in players' performances and fantasy returns, and decide the exact price to be set for each player for the upcoming football season.

I employed various clustering algorithms, including K-means clustering, Hierarchical Clustering, GMM Clustering, and DBSCAN clustering. Upon evaluating each cluster and their silhouette score, I identified the optimal clustering algorithm that best captured the inherent patterns within the data. These findings culminated in a visually appealing and comprehensible PowerPoint (PPT) presentation, crafted to highlight key conclusions and provide actionable recommendations.

Tool: Visual Studio Code, PPT (Presentation)
Tech Language: Python

View my Unsupervised Learning Project in Python on GitHub
View my PPT Presentation summarizing the Key Findings on GitHub