I am solving a problem of image classification of the image dataset for 3 classes. Dataset is highly imbalanced.
How will sampling (either over- or under-sampling) work in that case? Should I remove (or add) any random number of images, or should I follow some pattern?
In the case of CSV data, the general rule is to do PCA, and then remove the data points, but how to do it in the image dataset? Is there any other way to handle this problem?