Ensemble Based Approach: Evaluating network and Model Performance Tackling DDOS attack

Author:

Amit Dogra ORCID,Taqdir ORCID

Abstract

With in the dynamic realm of cyber threats, distributed denial of service (DDoS) attacks pose a serious threat. They can undermine network infrastructures and bring about service interruptions that cost money. Our research proposes an ensemble-based technique for DDoS attack detection in response to this problem. By combining the strengths of three distinct classifiers—Random Forest, K-Nearest Neighbors (KNN), and Adaboost—we create a powerful ensemble model. To ensure superior performance, we employ a Multi-Layer Perceptron (MLP) for intricate feature extraction and data normalization in the pre-processing stage. Together with individual classifiers, the ensemble's efficiency is carefully evaluated, verifying that it can accurately identify and counteract DDoS attacks. Motivated by the dynamic nature of DDoS attacks and their inability to be defended against by conventional defense mechanisms, our work is the first to apply machine learning to enhance detection. Ensemble approaches hold promise in addressing the evolving DDoS threat landscape because they combine multiple classifiers to enhance overall performance. The research adds a new dimension by combining MLP-based feature extraction with the Adaboost, KNN, and Random Forest classifiers to increase the discriminatory power of the model. Some of our objectives include building an ensemble-based DDoS attack detection system, evaluating individual classifier performance, comparing ensemble performance with individual classifiers, and using data normalization and MLP-based feature extraction. The research is methodically organized, with a literature review, methodology, performance analysis, ensemble approach analysis, and a concluding summary. The outcomes show the value of the recommended ensemble approach and pave the way for more advancements in DDoS attack detection methods, enhancing online service security and availability in the face of evolving cyber threats.

Publisher

Mallikarjuna Infosys

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3