A Comprehensive Study Of Classifying Imbalanced Data Methodologies

Authors

  • K.B. Jagadish Kumar, U. Mohan Srinivas, Dr. K. Navaz, M Akshitha Laasya, A. Ramesh

DOI:

https://doi.org/10.17762/msea.v68i1.1704

Abstract

In many real-world datasets, class instances are distributed unevenly. In a Class Imbalance Problem (CIP), certain class have a much larger amount of instances than the others. Due to inaccurate predictions of the poor class data samples, the unbalanced data reduces the accuracy of the prediction. Data mining experts from many different fields are familiar with CIP. A major challenge in machine learning (ML) and deep learning is how to classify data that is not evenly distributed (DL). Given its importance, the employment of sample techniques for increasing the classifier performance has attracted considerable attention in the current literature. In this part, we discuss the value of data organisation and the approaches used by numerous researchers to level the playing field in imbalanced classrooms. Various classifiers' accuracy and prediction rates have been analysed, and the criteria used to evaluate them have been examined...

Downloads

Published

2023-01-18

How to Cite

K.B. Jagadish Kumar, U. Mohan Srinivas, Dr. K. Navaz, M Akshitha Laasya, A. Ramesh. (2023). A Comprehensive Study Of Classifying Imbalanced Data Methodologies. Mathematical Statistician and Engineering Applications, 68(1), 76–89. https://doi.org/10.17762/msea.v68i1.1704

Issue

Section

Articles