CLASS IMBALANCE LEARNING METHODS FOR SUPPORT VECTOR MACHINES - Imbalanced Learning: Foundations, Algorithms, and Applications - page 81

CLASS IMBALANCE LEARNING
METHODS FOR SUPPORT
VECTOR MACHINES

RUKSHAN BATUWITA
National ICT Australia Ltd., Sydney, Australia
VASILE PALADE
Department of Computer Science, University of Oxford, Oxford, UK

Abstract: Support vector machines (SVMs) is a very popular machine learning
technique. Despite of all its theoretical and practical advantages, SVMs could pro-
duce suboptimal results with imbalanced datasets. That is, an SVM classifier trained
on an imbalanced dataset can produce suboptimal models that are biased toward the
majority class and have low performance on the minority class, as most of the other
classification paradigms. There have been various data preprocessing and algorith-
mic techniques proposed in the literature to alleviate this problem for SVMs. This
chapter aims to review these techniques.

5.1 INTRODUCTION

Support vector machines (SVMs) [1-7] is a popular machine learning technique,
which has been successfully applied to many real-world classification problems
from various domains. Owing to its theoretical and practical advantages, such
as solid mathematical background, high generalization capability, and ability to
find global and nonlinear classification solutions, SVMs have been very popular
among the machine learning and data-mining researchers.