wolffd@0: wolffd@0: wolffd@0: wolffd@0: Netlab Reference Manual gmminit wolffd@0: wolffd@0: wolffd@0: wolffd@0:

gmminit wolffd@0:

wolffd@0:

wolffd@0: Purpose wolffd@0:

wolffd@0: Initialises Gaussian mixture model from data wolffd@0: wolffd@0:

wolffd@0: Synopsis wolffd@0:

wolffd@0:

wolffd@0: 
wolffd@0: mix = gmminit(mix, x, options)
wolffd@0:

wolffd@0: wolffd@0: wolffd@0:

wolffd@0: Description wolffd@0:

wolffd@0: mix = gmminit(mix, x, options) uses a dataset x wolffd@0: to initialise the parameters of a Gaussian mixture wolffd@0: model defined by the data structure mix. The k-means algorithm wolffd@0: is used to determine the centres. The priors are computed from the wolffd@0: proportion of examples belonging to each cluster. wolffd@0: The covariance matrices are calculated as the sample covariance of the wolffd@0: points associated with (i.e. closest to) the corresponding centres. wolffd@0: For a mixture of PPCA model, the PPCA decomposition is calculated wolffd@0: for the points closest to a given centre. wolffd@0: This initialisation can be used as the starting point for training the wolffd@0: model using the EM algorithm. wolffd@0: wolffd@0:

wolffd@0: Example wolffd@0:

wolffd@0:

wolffd@0: 
wolffd@0: mix = gmm(3, 2);
wolffd@0: options = foptions;
wolffd@0: options(14) = 5;
wolffd@0: mix = gmminit(mix, data, options);
wolffd@0:

wolffd@0: wolffd@0: This code sets up a Gaussian mixture model with 3 centres in 2 dimensions, and wolffd@0: then initialises the parameters from the data set data with 5 iterations wolffd@0: of the k means algorithm. wolffd@0: wolffd@0:

wolffd@0: See Also wolffd@0:

wolffd@0: gmm

wolffd@0: Pages: wolffd@0: Index wolffd@0:

wolffd@0: