I saw in many tutorials in the Tensorflow 2 documentation, that they use two different functions to deal with categorical features:
- tf.keras.layers.experimental.preprocessing.StringLookup
- tf.feature_column.categorical_column_with_vocabulary_list
From what I've seen, they work very similarly and they are used for the same things... And I have also seen that categorical_column_with_vocabulary_list existed in Tensorflow 1, but StringLookUp didn't.
Is anyone experienced with these two functions and able to explain their advantages/disadvantages from each other?
Thank you very much.