Skip to content

Several state-of-the-art fuzzy clustering algorithms, including Fuzzy c-means clustering, fuzzy subspace clustering and maximum entropy clustering algorithms

License

Notifications You must be signed in to change notification settings

kailugaji/Fuzzy_Clustering_Algorithms

Repository files navigation

Fuzzy_Clustering_Algorithms

Several state-of-the-art fuzzy clustering algorithms, including Fuzzy c-means clustering, fuzzy subspace clustering and maximum entropy clustering algorithms.
Matlab code. Three examples in the iris data set.


Demo of FCM

FCM algorithm:

Run demo_fuzzy.m, choose the hyper-parameter 'choose_algorithm=1'. The clustering results:
Iteration 1, the number of iterations: 12, Accuary: 0.89333333
Iteration 2, the number of iterations: 12, Accuary: 0.89333333
Iteration 3, the number of iterations: 12, Accuary: 0.89333333
Iteration 4, the number of iterations: 12, Accuary: 0.89333333
Iteration 5, the number of iterations: 12, Accuary: 0.89333333
Iteration 6, the number of iterations: 12, Accuary: 0.89333333
Iteration 7, the number of iterations: 12, Accuary: 0.89333333
Iteration 8, the number of iterations: 12, Accuary: 0.89333333
Iteration 9, the number of iterations: 12, Accuary: 0.89333333
Iteration 10, the number of iterations: 12, Accuary: 0.89333333
The average iteration number of the algorithm is: 12.00
The average running time is: 0.01250
The average accuracy is: 0.89333333
The average rand index is: 0.87973154
The average normalized mutual information is: 0.74331694

Demo of FSC

FSC algorithm:

Run demo_fuzzy.m, choose the hyper-parameter 'choose_algorithm=2'. The clustering results:
Iteration 1, the number of iterations: 4, Accuary: 0.90666667
Iteration 2, the number of iterations: 4, Accuary: 0.90666667
Iteration 3, the number of iterations: 4, Accuary: 0.90666667
Iteration 4, the number of iterations: 4, Accuary: 0.90666667
Iteration 5, the number of iterations: 4, Accuary: 0.90666667
Iteration 6, the number of iterations: 4, Accuary: 0.90666667
Iteration 7, the number of iterations: 4, Accuary: 0.90666667
Iteration 8, the number of iterations: 4, Accuary: 0.90666667
Iteration 9, the number of iterations: 4, Accuary: 0.90666667
Iteration 10, the number of iterations: 4, Accuary: 0.90666667
The average iteration number of the algorithm is: 4.00
The average running time is: 0.00469
The average accuracy is: 0.90666667
The average rand index is: 0.89225951
The average normalized mutual information is: 0.80575367

Demo of MEC

MEC algorithm:

Run demo_fuzzy.m, choose the hyper-parameter 'choose_algorithm=3'. The clustering results:
Iteration 1, the number of iterations: 25, Accuary: 0.92000000
Iteration 2, the number of iterations: 25, Accuary: 0.92000000
Iteration 3, the number of iterations: 25, Accuary: 0.92000000
Iteration 4, the number of iterations: 25, Accuary: 0.92000000
Iteration 5, the number of iterations: 25, Accuary: 0.92000000
Iteration 6, the number of iterations: 25, Accuary: 0.92000000
Iteration 7, the number of iterations: 25, Accuary: 0.92000000
Iteration 8, the number of iterations: 25, Accuary: 0.92000000
Iteration 9, the number of iterations: 25, Accuary: 0.92000000
Iteration 10, the number of iterations: 25, Accuary: 0.92000000
The average iteration number of the algorithm is: 25.00
The average running time is: 0.00469
The average accuracy is: 0.92000000
The average rand index is: 0.90308725
The average normalized mutual information is: 0.75634802

Author of Code

Rongrong Wang (kailugaji)
My blog
2020/7/5

About

Several state-of-the-art fuzzy clustering algorithms, including Fuzzy c-means clustering, fuzzy subspace clustering and maximum entropy clustering algorithms

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages