Evaluating ML Algorithms for Network Intrusion Detection Insights from CICIDS2017 Dataset
Author(s):
K.RAMESH
Keywords:
CICIDS2017, Intrusion detection, Logistic Regression, Naive Bayes, and Decision Tree classifiers.
Abstract
The main aim of this study is to investigate the effectiveness of machine learning (ML) models in detecting network intrusions using the CICIDS2017 dataset. This dataset contains various instances of network traffic, encompassing different types of cyber-attacks. The study begins by consolidating and refining multiple datasets to create a unified dataset for analysis. Subsequently, an exploratory analysis reveals the distribution patterns of different attacks within the dataset. Data preparation involves optimizing the dataset for modeling by applying feature scaling and selection techniques. Several ML algorithms, such as Logistic Regression, Naive Bayes, and Decision Tree classifiers, are trained and rigorously evaluated using cross-validation methods. The evaluation metrics include accuracy measures and cross-validation mean scores, providing a comprehensive assessment of the models' performance. Moreover, the study employs the Random Forest classifier to identify and prioritize significant features aiding in intrusion detection. This research endeavors to contribute significantly to the field of cyber security by showcasing the potential of ML algorithms in detecting and categorizing diverse network intrusions. The findings highlight the feasibility of deploying robust ML-based intrusion detection systems, strengthening real-time network security applications and fortifying defenses against evolving cyber threats.
Article Details
Unique Paper ID: 163247
Publication Volume & Issue: Volume 10, Issue 11
Page(s): 1024 - 1030
Article Preview & Download
Share This Article
Join our RMS
Conference Alert
NCSEM 2024
National Conference on Sustainable Engineering and Management - 2024