Neural Model Options
Each model report has a red triangle menu that contains options for producing additional output or saving results. The model report red triangle menu contains the following options:
Diagram
Shows or hides a diagram that represents the hidden layer structure.
Show Estimates
Shows or hides a report of the parameter estimates.
Profiler
Shows or hides the prediction profiler. For nominal or ordinal responses, each response level is represented by a separate row in the Prediction Profiler. For more information about the options in the red triangle menu, see “Prediction Profiler Options” in Profilers.
Categorical Profiler
(Available only for nominal and ordinal responses.) Shows or hides a prediction profiler with all categorical responses combined into a single profiler row. For more information about the options in the red triangle menu, see “Prediction Profiler Options” in Profilers.
Contour Profiler
(Available only when the model contains more than one continuous factor.) Shows or hides the contour profiler, which shows the contours of the response graphically for two factors at a time. For more information about the options in the red triangle menu, see “Contour Profiler Platform Options” in Profilers.
Surface Profiler
(Available only when the model contains two or more continuous factors.) Shows or hides a three-dimensional surface plot. For more information about the options in the red triangle menu, see “Surface Plot Platform Options” in Profilers.
ROC Curve
(Available only for categorical responses.) Shows or hides the Receiver Operating Characteristic (ROC) plot that contains a curve for each level of the response variable. If you used validation, a plot is shown for each of the training, validation, and test sets. See “ROC Curve”.
Precision Recall Curve
(Available only for categorical responses.) Shows or hides the Precision-Recall Curve plot that contains a curve for each level of the response variable. A precision-recall curve plots the precision values against the recall values at a variety of thresholds. If you used validation, a plot is shown for each of the training, validation, and test sets. See “Precision-Recall Curve”.
Lift Curve
(Available only for categorical responses.) Shows or hides the Lift Curve plot. If you used validation, a plot is shown for each of the training, validation, and test sets. See “Lift Curve”.
Decision Threshold
(Available only for binary categorical responses.) Shows or hides Decision Thresholds reports for the training, validation, and test sets, if specified. Each report contains a graph of the distribution of fitted probabilities for each model, confusion matrices for each model, classification graphs to compare the model fits, and a table of classification accuracy metrics. See “Decision Thresholds Report”.
Plot Actual by Predicted
(Available only for continuous responses.) Shows or hides a plot with the actual values on the vertical axis and the predicted values on the horizontal axis.
Plot Residual by Predicted
(Available only for continuous responses.) Shows or hides a plot of the residuals on the vertical axis and the predicted values on the horizontal axis.
Save Formulas
Saves new formula columns to the data table. There are separate formula columns for the predicted response and the hidden layer nodes.
Save Profile Formulas
Saves a new formula column to the data table. The column contains a formula for the predicted response that includes embedded formulas for the hidden layer nodes. This option produces formulas that can be used by the interactive version of the profiler.
Save Fast Formulas
Save a new formula column to the data table. The column contains a formula for the predicted response that includes embedded formulas for the hidden layer nodes. This option produces formulas that evaluate faster than the other options, but cannot be used in the interactive version of the profiler.
Publish Prediction Formula
Creates prediction formulas and saves them as formula column scripts in the Formula Depot platform. If a Formula Depot report is not open, this option creates a Formula Depot report. See “Formula Depot”.
Make SAS Data Step
Creates SAS code that you can use to score a new data set.
Save Validation
Saves a new column in the data table. The column identifies which rows were used in the training and validation sets. This option is not available when a Validation column is specified on the Neural launch window. See The Neural Launch Window.
Save Transformed Covariates
Saves new formula columns to the data table. The new columns contain the formulas that are used to transform the covariates. This option is available only when the Transform Covariates option is checked on the Model Launch control panel. See Neural Fitting Options.
Remove Fit
Removes the entire model report.