Create an Inside Training Data Prediction

You can create a model that predicts whether an input is in the training data space [1] or out of it [0]. You can also export this model.

Create the inside training data prediction model

  • Extra menu > Create New Project for Inside Training Data Prediction.

  • Optionally, edit the min and max value of the inputs.

  • Optionally, edit the Number of Generated Data Points.

    This is the number of data points generated and marked as outside. A much higher number than the calculated heuristic can lead to bad results.

  • Click OK.

    The new ASCMO project opens.

  • Click the Inside Training Data button.

  • From the Model Type list, select a Classification model type.

    Optionally, edit the model properties. See, Model Types of ASCMO-STATIC.

  • Click OK.

  • Model training starts. When it is finished, you can visualize it as scatter plot (Data > Scatter Plot > Training Data) or 3D plot (In/Outputs > 3D Plot Outputs).

  • You can export the model using File > Export Model > *. See, Model Export .