| SOM Toolbox | Online documentation | http://www.cis.hut.fi/projects/somtoolbox/ |
[P] = cca(D, P, epochs, Mdist, alpha0, lambda0)
CCA Projects data vectors using Curvilinear Component Analysis.
P = cca(D, P, epochs, [Dist], [alpha0], [lambda0])
P = cca(D,2,10); % projects the given data to a plane
P = cca(D,pcaproj(D,2),5); % same, but with PCA initialization
P = cca(D, 2, 10, Dist); % same, but the given distance matrix is used
Input and output arguments ([]'s are optional):
D (matrix) the data matrix, size dlen x dim
(struct) data or map struct
P (scalar) output dimension
(matrix) size dlen x odim, the initial projection
epochs (scalar) training length
[Dist] (matrix) pairwise distance matrix, size dlen x dlen.
If the distances in the input space should
be calculated otherwise than as euclidian
distances, the distance from each vector
to each other vector can be given here,
size dlen x dlen. For example PDIST
function can be used to calculate the
distances: Dist = squareform(pdist(D,'mahal'));
[alpha0] (scalar) initial step size, 0.5 by default
[lambda0] (scalar) initial radius of influence, 3*max(std(D)) by default
P (matrix) size dlen x odim, the projections
Unknown values (NaN's) in the data: projections of vectors with
unknown components tend to drift towards the center of the
projection distribution. Projections of totally unknown vectors are
set to unknown (NaN).
See also SAMMON, PCAPROJ.