Train the plastics model

1. Open Models.
2. Select New model.
3. Name the model Plastics Unmixing.
4. Select Unmixing as the model type.
5. Select the Nylon and Belt dataset and its finalized version.
6. Confirm that both classes and the expected wavelengths and Reflectance processing level are shown.
7. Select Train.

Leave the default advanced settings for the first run unless you have a specific experimental reason to change them.

When training completes, inspect the validation metrics. This walkthrough's single-image split does not provide independent Test data. If you supplied untouched Test images, choose and lock the model version before generating the test run. Open Test Predictions to look for Belt responses on Nylon and Nylon responses on Belt. Do not expect a fixed score; see Interpreting model results.

If the model cannot be created or trained, confirm that the dataset version is finalized and contains at least two classes with valid training and validation data.

Next: Generate inference.