gmm properties

Gaussian Mixture node iconA Gaussian Mixture© model is a probabilistic model that assumes all the data points are generated from a mixture of a finite number of Gaussian distributions with unknown parameters. One can think of mixture models as generalizing k-means clustering to incorporate information about the covariance structure of the data as well as the centers of the latent Gaussians. The Gaussian Mixture node in SPSS Modeler exposes the core features and commonly used parameters of the Gaussian Mixture library. The node is implemented in Python.

Table 1. gmm properties
gmm properties Data type Property description
custom_fields boolean This option tells the node to use field information specified here instead of that given in any upstream Type node(s). After selecting this option, specify the fields below as required.
inputs field List of the field names for input.
target field One field name for target.
fast_build boolean Utilize multiple CPU cores to improve model building.
use_partition boolean Set to True or False to specify whether to use partitioned data. Default is False.
covariance_type string Specify Full, Tied, Diag, or Spherical to set the covariance type.
number_component integer Specify an integer for the number of mixture components. Minimum value is 1. Default value is 2.
component_lable boolean Specify True to set the cluster label to a string or False to set the cluster label to a number. Default is False.
label_prefix string If using a string cluster label, you can specify a prefix.
enable_random_seed boolean Specify True if you want to use a random seed. Default is False.
random_seed integer If using a random seed, specify an integer to be used for generating random samples.
tol Double Specify the convergence threshold. Default is 0.000.1.
max_iter integer Specify the maximum number of iterations to perform. Default is 100.
init_params string Set the initialization parameter to use. Options are Kmeans or Random.
warm_start boolean Specify True to use the solution of the last fitting as the initialization for the next call of fit. Default is False.