could anyone please give me an insight about what the ordinal loss defined in keras represents? https://www.tensorflow.org/ranking/api_docs/python/tfr/keras/losses/OrdinalLoss I'm not able to give an interpretation to that definition, as I don't understand for instance what is I_yi > j. In the documentation there is no citation of any artiecle. However, I'm looking for an objective fuction for image classification with ordinal categories, a sort of image ranking loss. Thanks Continue reading...
