Enhancing ML-Based DoS Attack Detection with Feature Engineering : IEEE CNS 23 Poster
- Resource Type
- Conference
- Authors
- Zhao, Shujun; Santana, Lesther; Owusu, Evans; Rahouti, Mohamed; Xiong, Kaiqi; Xin, Yufeng
- Source
- 2023 IEEE Conference on Communications and Network Security (CNS) Communications and Network Security (CNS), 2023 IEEE Conference on. :1-2 Oct, 2023
- Subject
- Communication, Networking and Broadcast Technologies
Computing and Processing
Statistical analysis
Telecommunication traffic
Machine learning
Network security
Feature extraction
Internet
Behavioral sciences
Denial of Service
feature selection
ML
- Language
Denial of Service (DoS) poses a serious threat to the Internet, causing significant financial harm. Modern DoS detection methods use statistical and machine learning (ML). They pick important features from network traffic for better model performance and accuracy. This study explores how feature selection improves ML-based DoS detection. It assesses feature importance using statistical analysis and feature engineering within DoS traffic datasets. Results show thorough statistical analysis and feature engineering help understand the DoS attack’s behavior and select optimal features for ML-based detection.