I am having pattern recognition class. My teacher need us to find out the difference between FCM and GMM, but I think these two clustering algorithm have no similarity. Hoping someone who knows could help me.
Asked
Active
Viewed 1,409 times
-1
-
Perhaps you can start by writing down the input and output of each algorithm more or less formally. You are then very likely to see what is similar about them. Figuring out the differences is the last stage... – Felix Goldberg Nov 29 '17 at 19:44
1 Answers
4
They have a lot in common, if you study them in more detail, and I'm pretty sure you can find these links discussed in various sources.
The main difference is that GMM has a nice and well understood theoretical model, assuming Gaussians and using maximum likelihood estimation, whereas FCM is using a very heuristic weighting approach, and you probably can't prove a lot about what it can and cannot do... I'd always prefer GMM to FCM.
Has QUIT--Anony-Mousse
- 76,138
- 12
- 138
- 194