In this post we'll be exploring a particular case when there's no labels when for a given example, i.e. it does not belong to any of the classes seen during training.
This notebook contains a series of experimentations using different approaches in the training process of a Convolutional Neural Network using a subsection of the MAMMOSET dataset