Generalized Linear Models Classification Model Viewer

This selection displays a successfully built Logistic Regression model, that is, a Generalized Linear Models (GLM) classification model, and its characteristics.

For a brief description of the Generalized Linear Models (GLM) algorithm, see Generalized Linear Models Algorithm.

Oracle Data Mining 11g uses GLM to solve two different kinds of problems: classification and regression. The resulting models have different viewers.

For a brief description of classification models, see Classification Models.

Menus

The following menus appear when the model is displayed in a new window:

Tabs

The model viewer has the following tabs:

Global Statistics

The Target Attribute for the model is displayed; this is the target selected when defining the Build Activity.

GLM classification models generate the following statistics that describe the model as a whole: Akaike's criterion for the fit of the intercept only model, Akaike's criterion for the fit of the intercept and the covariates (predictors) model, Schwarz's criterion for the fit of the intercept only model, Schwarz's criterion for the fit of the intercept and the covariates (predictors) model, -2 log likelihood of the intercept only model, -2 log likelihood of the model, likelihood ratio degrees of freedom, likelihood ratio chi-square probability value, pseudo R-square Cox and Snell, pseudo R-square Nagelkerke, dependent mean, percent of correct predictions, percent of incorrect predictions, percent of ties (probability for two cases is the same), number of parameters (the number of coefficients, including the intercept), number of rows, whether or not the model converged (yes or no), and whether or not a covariance matrix was computed (yes or no).

The following information is displayed for each metric:

To save the metrics information as a text file or a spreadsheet, click the Save As icon (a yellow arrow above a graphic of a page) above the grid that lists the metrics.

Coefficients

This tab displays the coefficients for a GLM classification (logistic regression) model.

The Target Attribute for the model is displayed; this is the target selected when defining the Build Activity.

The Target Class value is displayed.

You can select Show Intercept Row; the default is to not show the intercept row.

The coefficients are displayed in the Coefficients grid. If you want coefficients sorted by absolute value so that, for example, -0.5 is listed just before +0.5, click the Sort coefficients based on absolute value checkbox at the bottom of the window.

The default is to fetch (and display) 100 rows; to fetch the next 100 rows, click Refresh. To change the number of rows that are fetched, type a different number in the Fetch Size box and click Refresh. You can sort columns by clicking the column headings.

You can click Filter to filter the information displayed by including attributes, specifying maximum and/or minimum coefficient values, filtering by absolute value, and sorting in various ways. The default is to include all attributes, to not specify minimum and maximum values for coefficients, to not fileter by absolute value and to sort by attribute name in descending order.

The following information is displayed for each attribute:

To save the coefficients information as a text file or a spreadsheet, click the Save As icon (a yellow arrow above a graphic of a page) above the Coefficients grid.

Diagnostics

To see the Diagnostics tab, you must select Generate Diagnostics Table in the Build settings for the model. The default is to not generate a diagnostics table; if a diagnostics table is not generated, the Diagnostics tab is not displayed.

The following diagnostics are generated for logistic regression:

The diagnostics are displayed in a grid. The following control the display of diagnostics:

The following information is displayed in the grid:

To save the diagnostics as a text file or a spreadsheet, click the Save As icon (a yellow arrow above a graphic of a page) above the grid.

Results

The tab contains a tree view of test metrics and apply results. To view specific results, select the result, and click View. To delete a result, select one or more results and click Delete.

Build Settings

This tab displays the Type of the model (Classification), the Algorithm used to build the model, the Target Attribute of the model, Automatic Data Preparation (yes or no; see ADP for details), the Algorithm Settings used to build the model, and the Attributes used to build the model.

For detailed information about GLM settings for classification, see Generalized Linear Models Classification Settings.

The following algorithm settings are displayed:

Attributes used in the model build are displayed in a grid; the following information is displayed for each such attribute:

To save the attributes information as a text file, click the Save As icon (a yellow arrow above a graphic of a page) above the Attributes grid.

Row weights provide a weighting factor. To see normalized weights for target values, click Weights.

Task

This tab describes the task used to build the model. It shows when the task ran, what input was used, and what output was created.