Original Clustered
Neighbor points considered: More is more accurate but slower.
Gaussian sigma for scoring distances: At this distance, items get a score of 0.68.
Min Score for Merging: Higher minimum scores lead to more clusters.