CLUSTERING (dataset 2)

[TheExGenesis]

• KMeans varying the K

• Plot sillouette and rank index

• DBSCAN, determine distance metric (minkowski, euclidean, manhattan, chebyshev) with a base case of DBSCAN (sklearn default?) Keep in mind Hamming distance for categorical data.

• DBSCAN varying eps parameter with the chosen distance metric

• (Addressing statistical significance - p-val vs random data - might give points but oh well)