… The paper discusses aspects of classification applied to different data modalities,
including text, images, and audio. The classification challenges in the context of working
with imbalanced data are highlighted, the consequences of which can be mitigated by
applying data augmentation methods for underrepresented classes. Various classifiers,
from logistic regression to convolutional neural networks, can be used in classification
depending on several factors. The importance of evaluating model quality …