Feature classifier method?

Does the --p-optimize-feature-selection parameter only work for categorical data e.g. it helps improve accuracy when assigning genetic data to skin, mouth, gut like in the moving pictures tutorial? What features is this parameter selecting? - features from the reference database?

For my own notes, here's the documentation for that plugin:
https://docs.qiime2.org/2022.2/plugins/available/sample-classifier/classify-samples/

  --p-optimize-feature-selection / --p-no-optimize-feature-selection
        Automatically optimize input feature selection using
        recursive feature elimination.         [default: False]

This plugin is using features / taxa / ASVs / OTUs to classify samples. So the features being automatically removed here are the taxa in your table which are least predictive of sample type. Because it's removing features, it should not matter if your sample types are categorical or not.

Let us know if you have any other questions about this supervised learning stuff, or how to best use it with your data sets.