A Comparison between Support Vector Machines and Logistic Regression for Classification

Abstract

This dissertation is about classification methods for binary data developed in Computer Science and Statistics. The research focuses on two main algorithms called support vector machines and logistic regression. The thesis consists of three chapters. The first chapter provides a general discussion of classification algorithms used in Statistical and Machine Learning with special emphasis on logistic regression and support vector machines. The second chapter includes some simulation studies to compare the classification methods. The third chapter concludes the thesis with an application to a real dataset.