Gaussian Mixture node Build Options

Use the Build Options tab to specify build options for the Gaussian Mixture node, including basic options and advanced options. For details about these options not covered in this section, see the following online resources:

Basic

Covariance type. Select one of the following covariance matrices:
  • Full. Each component has its own general covariance matrix.
  • Tied. All components share the same general covariance matrix.
  • Diag. Each component has its own diagonal covariance matrix.
  • Spherical. Each component has its own single variance.

Number of components. Specify the number of mixture components to use when building the model.

Cluster Label. Specify whether the cluster label is a number or a string. If you choose String, specify a prefix for the cluster label (for example, the default prefix is cluster, which results in cluster labels such as cluster-1, cluster-2, etc.).

Random Seed. Select this option and click Generate to generate the seed used by the random number generator.

Advanced

Tolerance. Specify the convergence threshold. Default value is 0.001.

Number of iterations. Specify the maximum number of iterations to perform. Default value is 100.

Init parameters. Select the initialization parameter Kmeans (responsibilities are initialized using k-means) or Random (responsibilities are initialized randomly).

Warm start. If you select True, the solution of the last fitting will be used as the initialization for the next fitting. This can speed up convergence when fitting is called several times on similar problems.

The following table shows the relationship between the settings in the SPSS® Modeler Gaussian Mixture node dialog and the Python Gaussian Mixture library parameters.
Table 1. Node properties mapped to Python library parameters
SPSS Modeler setting Script name (property name) Gaussian Mixture parameter
Use predefined roles / Use custom field assignments role_use
Inputs predictors
Use partitioned data use_partition
Covariance type covariance_type covariance_type
Number of components number_component n_components
Cluster Label component_lable
Label Prefix label_prefix
Set random seed enable_random_seed
Random Seed random_seed random_state
Tolerance tol tol
Number of iterations max_iter max_iter
Init parameters init_params init_params
Warm start warm_start warm_start

1 Scikit-learn: Machine Learning in Python, Pedregosa et al., JMLR 12, pp. 2825-2830, 2011.

2 "User Guide." Gaussian mixture models. Web. © 2007 - 2017. scikit-learn developers.