Evaluation of dataset imbalance and its infuence together with other characters of the dataset to machine learning quality

Abstract

A qualitative dataset is one of the important components for all types of machine learning. However, the vast majority of real datasets are imbalanced. Despite the large number of methods that can work with imbalanced data and approaches to using these methods, imbalance in most cases poses a problem for the work of machine learning methods.