Please ensure Javascript is enabled for purposes of website accessibility
Powered by Zoomin Software. For more details please contactZoomin

AVEVA™ Unified Engineering

Machine Learning functionality

  • Last UpdatedFeb 04, 2026
  • 3 minute read

Take the following steps to build an ML model:

  1. Create machine learning (ML functionality)

    • On the Admin tab, in AI group, select PreDAI to display the Predictive Design AI window.

    • Select an ML Model World from the dropdown list.

    • Enter the ML model name.

    • Enter the ML model description.

    • Select Refresh to update the list of ML Model Worlds in the dropdown list.

    • Select View Existing Model to view or delete models.

  2. Input Data for ML Model

    • Import the CSV file with the dataset.

    • The DataTypeForm window displays a list of columns - Name, Label and Datatype.

    • Select a Purpose.

    • Select a Datatype, which is optional.

    • Select Apply.

      Note:
      The CSV header must be alphanumeric. It must start with an alphabet character or underscore (_)
      The Purpose option displays Feature by default.

  3. Select ML task

    • Select the model type: Classification or Regression.

  4. Select training/testing split ratio

    Select training/testing split ratio determines what percentage of the dataset is used to train the model and what percentage is for testing.

    • Adjust the trackbar for Select training/testing split ratio.

      Note:
      The training percentage must be set between 50% and 90%.

  5. Select training algorithm

    The Auto-select the Algorithm enables the application to check the algorithms and automatically train the model with the highest accuracy.

    Clear the Auto-select the Algorithm checkbox to select an algorithm from the dropdown list and train the model using the selection.

    Each algorithm has its own set of hyperparameters that you can customize.

    • Select the algorithm from the dropdown list or Auto select the algorithm.

      Note:
      If the Auto-select the algorithm checkbox is cleared, the algorithm dropdown list becomes active.

    • The Hyperparameters Settings button is activated, allowing for the adjustment of hyperparameters.

  6. Select Build ML Model to display the Real-Time Log Form.

TitleResults for “How to create a CRG?”Also Available in