<![CDATA[Handling imbalanced data is an important issue that can affect the validity and reliability of the results. One common approach to addressing this issue is through re-sampling the data. Re-sampling is a technique that allows researchers to balance the class distribution of their dataset by either over-sampling the minority class or under-sampling the majority class. Over-sampling involves adding more copies of the minority class examples to the dataset in order to balance out the class distribution. On the other hand, under-sampling involves removing some of the majority class examples from the dataset in order to balance out the class distribution. It's also common to combine both techniques, usually called hybrid sampling. It is important to note that re-sampling techniques can have an impact on the model's performance, and it is essential to evaluate the model using different evaluation metrics and to consider other techniques such as cost-sensitive learning and anomaly detection. In addition, it is important to keep in mind that increasing the sample size is always a good idea to improve the performance of the model. In this systematic review, we aim to provide an overview of existing methods for re-sampling imbalanced data. We will focus on methods that have been proposed in the literature and evaluate their effectiveness through a thorough examination of experimental results. The goal of this review is to provide practitioners with a comprehensive understanding of the different re-sampling methods available, as well as their strengths and weaknesses, to help them make informed decisions when dealing with imbalanced data.]]>
Copyrights © 2023
