PCA/Factor Node Expert Options

If you have detailed knowledge of factor analysis and PCA, expert options enable you to fine-tune the training process. To access expert options, set Mode to Expert on the Expert tab.

Missing values. By default, IBM® SPSS® Modeler only uses records that have valid values for all fields used in the model. (This is sometimes called listwise deletion of missing values.) If you have a lot of missing data, you may find that this approach eliminates too many records, leaving you without enough data to generate a good model. In such cases, you can deselect the Only use complete records option. IBM SPSS Modeler then attempts to use as much information as possible to estimate the model, including records where some of the fields have missing values. (This is sometimes called pairwise deletion of missing values.) However, in some situations, using incomplete records in this manner can lead to computational problems in estimating the model.

Fields. Specify whether to use the correlation matrix (the default) or the covariance matrix of the input fields in estimating the model.

Maximum iterations for convergence. Specify the maximum number of iterations for estimating the model.

Extract factors. There are two ways to select the number of factors to extract from the input fields.

• Eigenvalues over. This option will retain all factors or components with eigenvalues larger than the specified criterion. Eigenvalues measure the ability of each factor or component to summarize variance in the set of input fields. The model will retain all factors or components with eigenvalues greater than the specified value when using the correlation matrix. When using the covariance matrix, the criterion is the specified value times the mean eigenvalue. That scaling gives this option a similar meaning for both types of matrix.
• Maximum number. This option will retain the specified number of factors or components in descending order of eigenvalues. In other words, the factors or components corresponding to the n highest eigenvalues are retained, where n is the specified criterion. The default extraction criterion is five factors/components.

Component/factor matrix format. These options control the format of the factor matrix (or component matrix for PCA models).

• Sort values. If this option is selected, factor loadings in the model output will be sorted numerically.
• Hide values below. If this option is selected, scores below the specified threshold will be hidden in the matrix to make it easier to see the pattern in the matrix.

Rotation. These options enable you to control the rotation method for the model. See the topic PCA/Factor Node Rotation Options for more information.