I'm trying to do some fuzzy clustering in spatial data. Now I'm looking for the best indicator of Fuzzy Clustering validity in Spatial data. I have read about PC, CE, SC, S, XB, DI, ADI, and IFV indices. Anyone have an idea?
XB can be suitable for the evaluation of fuzzy clustering because It considers the fuzziness when evaluating the clusters.
Recall, precision and F-score are suitables for the evaluation of fuzzy clustering.
Hi,
When I implemented Fuzzy Clustering, I used Fuzzy Joint Points. This kind of algorithm is pretty good because you can use the mathematical proposal as validation index. If you want to know more about this algorithm, refer to:
E. N. Nasibov & G. Ulutagay. A new unsupervised approach for fuzzy clustering. Fuzzy Sets and Systems, 158(19):2118–2133, 2007. ISSN 0165-0114.
Bests,
Wilfredo
Hi Anisur and Mustapha, thank you for your answer. Actually I'm trying to use it in my spatial data clustering, which is considering geographical attributes. Some literatures said that IFV is better. But I still don't get sufficient reason. Any idea?
Hi Wilfredo, thank you so much it sounds good, I will try to collect more information about this. Actually my simulation is implemented in Matlab, so if the index code is available, it would be very helpfull.
- Badges
- Science topic
- Similar topics
- Computer Science
- Data Mining