Agglomerative hierarchical clustering algorithm

Hierarchical clustering algorithms create a hiearchical decomposition of the data set using some criterion. The agglomerative (bottom-up) method merges clusters iteratively.

  • Place each object in its own cluster
  • Merge these atomic clusters into larger and larger clusters
  • Continue until all objects are in a single cluster
  • Most hierarchical methods belong in this category; they differ only in their definition of between-cluster similarity.

AGNES (Agglomerative Nesting)

Use the Single-Link method and a dissimilarity matrix. We merge nodes that have the least dissimilarity and go on in non-descending fashion. Eventually all nodes belong to the same cluster.


Practicality

This could be used to segment MRI data into 'like' clusters. Another algorithm could then be used to determine a tissue-labeling for the 'like' clusters individually.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License