Dear researchers, I'm working for my thesis about portfolio. I decided to use FCM as a clustering technique to group the stock data in few clusters then use Genetic Algorithm to optimize it. I'm not sure if Partition index and partition coefficient are enough to determine the number of cluster for FCM. I need some of your professional opinions and ideas.
Dear Charlene,
You have selected a very interesting research project please check my publications which is related to your question. One of these publications is using FCM and hybrid GA. Another is similar to what you are trying to do but it uses SOM and GA. Please let me know if you need a copy best wishes.
Hi,
I think that you can use another clusters validity indices such as WB or the newest one called STR index with little modification (Here the title of the article related to STR: A new validity index for crisp clusters - Artur Starczewski).
Good Luck
good day,
If you don't mind Mr. M. Awad, would you send me those copies of yours. Thank you
Mr. Habib Mahi, I googled the article but not downloadable. If you have a copy of it, would you please send me a copy (A new validity index for crisp clusters - Artur Starczewski). Thank you,
Please read my chapter:
Soft Clustering: Algorithms and Validity of Solutions
Fuzzy Computing, pp. 445-472, 1988.
It has a better approach and nine validity measures.
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress