Home > SFMI > InputClustering1D.m

InputClustering1D

PURPOSE ^

Perform Fuzzy C-Mean Clustering in each indata dimension separately

SYNOPSIS ^

function [input,LastMfNumbers]=InputClustering1D(xtrain,indata,LastMfNumbers,options)

DESCRIPTION ^

 Perform Fuzzy C-Mean Clustering in each indata dimension separately 
 (1D clusterings)
-------------------------------------------------------------------------
 Parameters
-------------------------------------------------------------------------
 xtrain - matrix containing the training data points
        Each row corresponds to a data point, each column corresponds to a
        dimension.
 indata(i) --name
          |-range
 LastMfNumbers - row vector containing the ordinal numbers of the
                 membership functions for each dimension
 options - column vector containing parameters of the FCM clustering
   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: 0)
   options(5):  FS index plotting (default: 0)
   options(6):  indicates whether cluster merging should be done (default: 1)
   options(7):  indicates whether the clusters should be plotted (default: 0)
   options(8):  indicates whether the approximated trapezoids should be
                plotted (default: 0)
   options(9):  indicates whether the number of clusters can be reduced to
                1 during cluster merging (default: 0)
   options(10): similarity treshold for set merging (default: 0.05)
   options(11): fuzzy index for the calculation of the FS index 
                (default: 2)
   options(12): cut level for the approximation of the trapezoids 
                (default: 0.85)
   options(13): info display during optimal cluster number determination 
                (default: 0)
   options(14): maximum number of clusters (default: 3)
   options(15): number of decimals (default: 2)
   options(16): initialization mode of the membership matrix by
                FCM (default: 0('deterministic'))
   options(17): treshold for the examination of the identity of two
                cluster centers (default:1e-6)
 -------------------------------------------------------------------------
 Returned values
 -------------------------------------------------------------------------
 Vector containing the structures that decribe linguistic variables
 input(i) --name
          |-range
          |-mf(j)--name
                 |-RPx
 -------------------------------------------------------------------------
 Remarks
 -------------------------------------------------------------------------
 
 
 Zsolt Csaba Johanyák, johanyak.csaba@gamf.kefo.hu, v. 1.2,  23. February 2007.

CROSS-REFERENCE INFORMATION ^

This function calls: This function is called by:
Generated on Thu 26-Mar-2009 13:11:50 by m2html © 2005