The
Lasso Regression results dialog box is displayed after the coordinate descent is completed. The name of this dialog box is either
Linear Results or
Logistic Results, according to the algorithm selected on the
Quick tab of the
Lasso Regression dialog box. The results dialog box contains options to review the details of the final solution in diagnostic results tables and graphs, and for saving deployment information for the chosen model in PMML script.
The following options are available regardless of which tab is selected:
| Element Name
|
Description
|
| Summary
|
Click this button to produce the summary statistics for the respective algorithm’s analysis. These values will be identical to the values displayed in the
Summary box located at the top of the
Lasso Regression results dialog box.
|
| Cancel
|
Click this button to return to the
Lasso Regression dialog box.
|
| Options
|
Click this button to display the
Options menu.
|
| Code generator
|
If your program is licensed to include this feature, you can generate computer code to implement the current model for predicting new observations. When you click this button, you have the following choice:
|
| PMML script
|
This command will generate code in Predictive Model Markup Language (PMML) which is an XML-based language for storing information about fully trained (parameterized) models and for sharing those models with other applications.
Note: PMML script is produced for the particular model specified by the lambda value under
Model results group. Please note the categorical variable levels are coded using indicator coding in the analysis but presented using over-parametrized coding in the script. To obtain accurate predictions, it is suggested that the users run the chosen model in Statistica General Linear Model(Linear Regression) or Statistica Generalized Linear Model(Logistic Regression) and use the deployment option available there.
|
