Pitchers are clustered based on over 50 pitch metrics. Hitter's statistics are compared vs each cluster of pitchers. Bayesian data analysis provides insights for how better or worse a player is vs each cluster.
Scrape RotoWire for each hitter vs pitcher matchup statistics. Loop through all active players on team, write the scraped data to json files.
View Script
Clean data to handle Null Values
View Notebook
Scale data to account for large value differences between predictors
View Notebook
Fit the KMeans Clustering model on training data. Evaluate its accuracy
Organize and aggregate data in the python for easier interpretation
Deploy the Streamlit app
