After the ART2 neural network determines the number of clusters,
standard clustering algorithms can partition this space into
groups of sessions that are close to each other based on a distance
measure. The well-known K-means algorithm is used as the base
method for clustering interest-based representation sessions and
sequence-based representation sessions. The K-means clustering
algorithm groups sessions by attributes/features into a k (positive
integer) number of groups by minimizing the sum of squares of
distances between data and the corresponding cluster centroid.
Additionally, the most popular Euclidean distance is used as the
distance measure. The K-means clustering algorithm is performed
in the following steps: Step 1: Generate initial random cluster centroids
for k clusters and k obtained by ART2 neural network. Step 2:
Assign each session to its closest cluster centroid in terms of
Euclidean distance. Step 3: Compute new cluster centroids. Step
4: If cluster memberships differ from the last iteration, repeat steps
2–3. Step 5: Stop and store clustering result.