A classifier is trained for a specific subset of features. If the feature set used is forgotten, then the classifier itself is useless. Therefore the feature manager object should be saved along with the classifier. Information about the number of channels is also necessary.
A classifier is trained for a specific subset of features. If the feature set used is forgotten, then the classifier itself is useless. Therefore the feature manager object should be saved along with the classifier. Information about the number of channels is also necessary.