This repository contains my solutions for a lecture about fuzzy clustering. No guarantee for correctness or anything, but maybe someone finds the code useful anyway.

Content:

• Fuzzy C-Means (FCM)
• A benchmark of various different methods to calculate covariance matrices
• Various cluster validity indices
  • Partition Coefficient (PC)
  • Normalized Partition Coefficient (NPC)
  • Fuzzy Hypervolume (FHV)
  • Fukuyama-Sugeno Index (FS)
  • Xie-Beni Index (XB)
  • Beringer-Hullermeier Index (BH)
  • Bouguessa-Wang-Sun index (BWS)

Here is a plot of some data clustered with FCM:

And here is a plot of the various cluster validity indices for that data clustered with different number of clusters: