While K-means clustering assigns data points based on their proximity to a centroid, DBScan examines the concentration of data points. A cluster in DBScan is defined as a region with a high concentration of data points surrounded by a region of low concentration. The parameter input into DBScan is epsilon which is a measure of the distance a data point may be from the cluster and still considered part of the cluster.

The workflow for this portion of the project draws heavily from the following:

https://scikit-learn.org/stable/auto_examples/cluster/plot_dbscan.html#sphx-glr-auto-examples-cluster-plot-dbscan-py

https://www.machinecurve.com/index.php/2020/12/09/performing-dbscan-clustering-with-python-and-scikit-learn/

https://www.coryjmaklin.com/machine-learning-clustering-dbscan-determine-the-optimal-value-for-epsilon-eps-python-example

The following values for epsilon were determined by computing the point of greatest curvature of the graph of distances for a particular dataset.

Distance versus index value

Based on DBScan each of the datasets contain only one cluster with a few outliers.