Banks financial data: cluster results

50 items have been grouped optimally into 3 clusters based on 12 columns using K-Means.

This cluster has a score of 16. (Larger values indicate better clusters. There's no "minimum" value.)

Each cluster has different characteristics. The average features for every cluster are in this table.
Adjust the slider below to highlight important features.

This plots the clusters on every pair of features. Use it to get a sense of how the clusters are different from each other.
Click anywhere to zoom out or in.

Every item's cluster and features are shown below. You can sort the table and explore the clusters.

Some entities are more similar (and dissimilar) than the rest as determined by the Euclidean distance.
Displays 50 similar and dissimilar entities. Each entry in the entity-pair corresponds to a data item.

Clustering Application