Parameter Range Automated Machine Learning (ASCMO-STATIC)

Model menu > Automated Machine Learning > Parameter Range button

EIn the Automated Machine Learning window you can specify the settings for the automated machine learning and the range of hyperparameters. In the Parameter Range Automated Machine Learning window there are separate tabs for the MLP and ASC-General options.

The Parameter Range Automated Machine Learning window contains the following elements:

    Modeling Method

    Activate the checkboxes for the modeling methods you want to use for automated machine learning.

    Automatic Input Selection

    Automatically learn to base the model on only the important inputs, as a trade-off between model quality and complexity. The global settings are taken into account and each model input list is a subset of the globally activated inputs.

    This option is ignored if only one global input is selected.

    Random Seed

    Enter a seed value for the random number generator to ensure reproducible results.

Validation Data

Choose whether the selection of the best model (RMSE) is based on the training or test data, if available.

Output Transformation

Select the transformation type of the output. Using a transformation can improve the model prediction. Not all transformations are available if the training data has negative or zero values.

You can select from the following choices:

  • none: no transformation
  • 1/y: inversion
  • 1/sqrt(y): inverted square root
  • log(y): logarithm
  • sqrt(y): square root

  • Bounded: limited to lower and upper bound

  • log(y+c): logarithm plus constant

 

ClosedModel Setting: MLP

    Activate the MLP checkbox on the General tab if you want to include this element in the machine learning process to find the best result. The more checkboxes you select, the longer it takes.

    Activate the checkboxes of the elements to be used during the machine learning process:

    • Activation Function : Activate the corresponding activation functions for the hidden layers.

    • Output Function : Activate the corresponding output functions.

    Enter the range of minimum and maximum values for the continuous parameters to be used during the automated machine learning process:

    • Number of Neurons : Enter the range of neurons per layer.

    • Number of Layers: Enter the range of hidden layers.

    • Number of Iterations: Enter the range of iterations.

ClosedModel Setting: ASC-General

    Activate the ASC GP checkbox on the General tab if you want to include this element in the machine learning process to find the best result. The more checkboxes you select, the longer it takes.

    Activate the checkboxes of the elements to be used during the machine learning process:

    • Validation Mode: Activate the checkbox if you want Leave-One-Out error on training data to be used to calculate the RMSE. If activated, the Validation Data setting in the General tab is ignored.

    • Kernel: Activate the checkboxes of the corresponding model kernels to be used.

    Enter the range of minimum and maximum values for the continuous parameters to be used during the automated machine learning process:

    • Number of Iterations: Enter the range of iterations.

    • Subset Size: Enter the range of the maximum number of basic functions for the model training.

Default

Sets all parameters to their default values.

OK

Applies your settings and closes the window.

Cancel

Discards your settings and closes the window.

See also

Automated Machine Learning window