I am having a dataset with highly imbalanced classes (some classes have 4k examples while others have only one example). What is the best approach to handle such problem?
Traditional approaches are oversampling, undersampling and weight in loss function but to me it looks that all this approaches will totally miss the minority class. All this approaches works in a situation where there is several hundreds examples from the minority class.
Also, using GAN to learn generative model to generate new examples just for single example per class in training set is not feasible.
