Use k-means clustering on playlists to find patterns

[jbrown1618]

Once the clusters are identified, we would use p-values from hypothesis tests to identify the features that differentiate the clusters the most