Model Configurations (ASCMO-DYNAMIC)

The Model Configurations window Model > Configurations provides a separate tab for each output. Each tab contains elements to determine the model properties individually for each output. You can define several configurations in which you specify the modeling method and the model type.

Note
For an overview of all model types and their best use, see Overview: Model Type Descriptions.

Output Name

You can use this field to rename the output.

Output Unit

You can use this field to change the unit of the output.

Model Configuration

Select an already created configuration for editing and applying or create a <new> configuration to apply to the output. Enter configuration name in the text field above the button row.

Modeling Method

Select the modeling method you want to use.

Model properties area

The settings available in the model properties area depend on the selected modeling method.

No Model (initial state)
NARX Structure with optional Automatic Feature Selection
Recurrent Neural Network
Convolutional Neural Network
Static Model
Ensemble Model
Anomaly Detection (PCA)
Anomaly Detection (Autoencoder)

Configuration Name

Insert a name for the configuration .

Default

Sets all parameters to their default values.

Use Auto ML

Opens the Automated Machine Learning to perform automated machine learning to find properties (Network Layout, Output Properties, Training Properties) automatically.

Export Job to M Script

Exports the current model settings to a MATLAB Script file (*.m) to outsource the training (e.g. in MATLAB® on a server).

Use for further options:

Export Job to Docker: Exports the single result optimization information as *.docker.ascmo file. Use the file to perform the optimization in a Docker container, e.g., in the cloud.

Opens the online help in the appropriate context.

Starts the model training and closes the window.

Apply

Starts the model training without closing the window.

Cancel

Discards your settings and closes the window.