<< snorm sciFLT tnorm >>

sciFLT >> sciFLT > subclust

subclust

Find cluster centers with substractive clustering.

Calling Sequence

[centers,sigmas]=subclust(X [,r [, opt ]])

Parameters

X:

matrix of reals.The pairs of inputs points.

r:

real vector of real.

opt:

[1x3] vector of real.

centers:

integer, maximum number of iterations. The defaul value is 100

sigmas:

scalar, minimum change value between two consecutive iterations. The default value is 0.001

Description

subclust find cluster centers with substractive clustering for a matrix X in which each row contains a data point. r is a vector that specifies a cluster center's range of influence in each of the data dimensions, assuming the data falls within a unit hyperbox (internalli relized).

opt=[ACCEPT_RATIO, REJECT_RATIO, REDUCTION_RATIO]. If opt is not provided, then the routine use the followings: ACCEPT_RATIO=0.5, REJECT_RATIO=0.15, REDUCTION_RATIO=1.5.

Examples

// Take 400 random pairs of points centered at (0,0), (3,3), (7,7) and (10,10)
Xin=[rand(100,2);3+rand(100,2);7+rand(100,2);10+rand(100,2)];
// Find clusters
centers=subclust(Xin);

See also

Authors


Report an issue
<< snorm sciFLT tnorm >>