A Detailed Analysis of the Multi-Class Classification Problem in Network Intrusion Detection using Resampling Techniques

It appears your Web browser is not configured to display PDF files. Download adobe Acrobat or click here to download the PDF file.

Click here to download the PDF file.

Creator: 

Ahsan, Rahbar

Date: 

2021

Abstract: 

Machine learning algorithms are known to help identify cyberattacks such as network intrusion. However, common network intrusion datasets are often imbalanced. We conduct a detailed analysis on the impact of the different resampling techniques over different machine learning classifiers. We include more advanced resampling techniques, such as CGAN oversampling, in our study and compare its performance against other oversampling techniques. To further investigate CGAN-based oversampling potential, we examine the effect of CGAN with other standard machine learning classifiers on two different datasets. We do not recommend using CGAN in a dataset with extremely low samples in its minority classes based on our experimental results. Consequently, we also investigate the choice of minority class(es) to be oversampled in a dataset with low minority samples. Finally, the impact of the number of synthetic samples to be generated on the detection rate is evaluated on two different network intrusion datasets.

Subject: 

Computer Science

Language: 

English

Publisher: 

Carleton University

Thesis Degree Name: 

Master of Computer Science: 
M.C.S.

Thesis Degree Level: 

Master's

Thesis Degree Discipline: 

Computer Science

Parent Collection: 

Theses and Dissertations

Items in CURVE are protected by copyright, with all rights reserved, unless otherwise indicated. They are made available with permission from the author(s).