MATLAB: Set the number of iterations in FCM algorith

clusteringfcmfuzzyFuzzy Logic ToolboxMATLAB

Hello,

The general form of fuzzy C-means is

[center,U,obj_fcn] = fcm(data,cluster_n,options)

options(1): exponent for the partition matrix U (default: 2.0)

options(2): maximum number of iterations (default: 100)

options(3): minimum amount of improvement (default: 1e-5)

options(4): info display during iteration (default: 1)

How can I change the number of iterations to 500?

I get the message "too many input arguments", if I use

fcm("my_data",10,2,500)

here 10 is the number of clusters and 2 is the exponent

Thank you.

Pavlos

Best Answer

fcm("my_data",10,[2,500])

Related Solutions

MATLAB: How to view all U matrix for each iteration in Fuzzy c means clustering

Loop around from 1 to the number of iterations you want. The default for fcm() is 100; if you wanted to match that, then you would loop to 100. Call the loop control variable L

In each loop, you would set the random number generator seed to the same value. Then call fcm() passing in your data, then the number of clusters, and then the vector [2, L, 1e-5, 1] which will be your options vector. The second element of the options vector is the number of loop iterations that fcm is to do, so each time you will be telling fcm to do one more iteration than the time you ran it before. Record or otherwise process the U result.

For example:

randseed = 54321;
numcluster = 5;
maxiter = 100;
U = cell(maxiter, 1);
for L = 1 : maxiter
  rng(randseed);   %must be reset to the same value each time
  [~, U{L}, ~] = fcm(YourData, numcluster, [2, L, 1e-5, 1]);
end

Now U{1} will be the U after the first iteration, U{2} will be the U after the second iteration, and so on.

MATLAB: How to get fuzzy-ELA algorithms for deinterlacing

fuzzy logic

Best Answer

Related Solutions

MATLAB: How to view all U matrix for each iteration in Fuzzy c means clustering

MATLAB: How to get fuzzy-ELA algorithms for deinterlacing

Related Question