-
From the visualization controls in the upper right of the Visual Analytics
canvas, click New Visualization
 .
-
Select Approximation Measure.
-
Enter a Title and
Description.
The title appears on the visualization when it is added to the canvas.
-
Select one or more Approximations.
-
Select an Output parameter.
- Optional:
Select
Show Actual vs Predicted.
A scatter plot of actual parameter data versus predicted parameter data
based on the selected approximation appears in the visualization.
- Optional:
Select Show Residual vs Predicted.
The residual plot shows residuals for each output parameter value on the
Y-axis and predicted values for the output parameter, based on the selected
approximation, on the X-axis. The residual plot appears in the
visualization.
- Optional:
Select Show Cross Validation % Error.
A histogram showing the error for each degree of the approximation
polynomial appears in the visualization.
-
Select a Measure:
| Option | Description |
|---|
| Validation |
Lists fit and error measures for the validation data set used for
the approximation. |
| Training |
Lists fit and error measures for the data set used to train the
approximation. |
For more information, see Measure of Fit.
-
Do one of the following:
- Click Create and Close to create the new
visualization and to return to the canvas.
- Click Create to create the new visualization and
to keep the new visualization window open.
|