next up previous
Next: About this document Up: Partitional Clustering Previous: Nearest Neighbor clustering

Fuzzy Clustering

In this type of clustering the objects are allowed to belong to more then one cluster and has a degree of membership assigned from [0,1].
Algorithm

  1. Select an initial partition. Repeat steps 2 to 4 until the cluster memberships stabilize.
  2. Computer the membership functions.
  3. Compute the criterion function.
  4. Reclassify patterns to improve the value of the criterion function.
Fuzzy clustering is useful in cases where the clusters are touching or overlapping.

Miranda Maria Irene
Thu Apr 1 15:43:18 IST 1999