[Index for export/lyngby]
lyngby_ui_km_init - GUI for K-means parameter initialization
Global: (from lyngby_km_global)
This function controls the window for choosing the parameters
used in the K-means clustering. In the window the following
parameters can be set:
- Type: Whether the median or the mean of the cluster members
is used to determine the next iterative cluster
- Standardization: How to standardize (normalize) the
individual objects (voxels). The objects can be
standardize with their standard deviation (sigma) or
with their range (the difference between the maximum
and the minimum).
- Clusters: Number of clusters. If the number of clusters are
not known in advance you can use the iterative
K-means clustering. See lyngby_ikm_main
- Initial cluster centers: Determines how the cluster centers
should be initialized.
- Decay rate: (between 0 and 1) The weight between the cluster
center in the previous iteration and the new estimate
of the cluster center. (This will regularize the
performance). A value of zero will only use new
- Iterations (>0): Maximum number of iterations. If this is
too small the optimization of the cluster centers
might not be complete.
- Variable: Whether to cluster on the raw data vectors or the
cross correlation between the paradigm and data
- Components: Number of cross-correlation components to
cluster from. This only applies when the variable
(the metric) is cross-correlation.
When you have changed the parameters the 'Apply' button must
be pressed before it will have an effect.
See also LYNGBY, LYNGBY_KM_MAIN, LYNGBY_KM_GLOBAL,
LYNGBY_UI_MAIN, LYNGBY_IKM_MAIN, LYNGBY_UI_IKM_INIT.
$Id: lyngby_ui_km_init.m,v 1.17 2002/09/27 12:51:28 fnielsen Exp $
Cross-Reference InformationThis function calls
This function is called by
Produced by mat2html on Wed Jul 29 15:43:40 2009
Cross-Directory links are: OFF