Intelligent Network Intrusion Detection System Using Machine Learning and deep Learning

DOI

10.2991/978-94-6463-858-5_52How to use a DOI?

Keywords

Malware Detection; Random Forest; Convolution Neural Network; preprocessing; Feature Selection; Label Encoding

Abstract

Over the last few years, the demand for effective and precise malware and intrusion detection systems has been on the rise because of mounting cybersecurity threats. Conventional signature-based methods tend to face a high false positive rate, but machine learning models offer a robust alternative through feature selection and classification methods. We present, in this paper, Random Forest-based classification model for selecting the highest-ranked significant features to enhance detection accuracy and CNN (Convolution Neural Network) is employed for spatial feature extraction. It extracts local patterns and dependencies from the network traffic data, including packet structure and flow patterns. CNN layers apply a filter to the input data, deriving high-level spatial features that are of relevance to intrusion detection. The dataset contains numeric and categorical features, with categorical variables encoded with Label Encoding and for numeric data. Min-Max normalization is performed before feature selection is applied. The Random Forest algorithm provides ranked feature importance, and the top 10 most important features are used for training the model. A Random Forest classifier with 100 trees is trained on the refined dataset, achieving high training accuracy (~100%), while the test accuracy is slightly lower, indicating possible overfitting. The proposed model undergoes five key stages: data preprocessing, feature selection, model training, evaluation, and virus detection in test samples. The training process is computationally efficient, requiring only a few seconds. The evaluation metrics, including classification reports and accuracy scores, highlight the model’s effectiveness in malware and intrusion detection. The results demonstrate that Random Forest’s feature selection capability enhances classification performance, making it a powerful tool for cybersecurity applications. The multi-classification model trained on the NSL-KDD + dataset demonstrates outstanding performance, the model maintains exceptional precision (98.25%), reaching a peak accuracy of 99.08%.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - V. Krishna Sameera
AU  - Kotcharla Sravika
AU  - Koulury Shiva
AU  - Ruthala Gayathri
AU  - Duvvapu Sadwika
PY  - 2025
DA  - 2025/11/04
TI  - Intelligent Network Intrusion Detection System Using Machine Learning and deep Learning
BT  - Proceedings of International Conference on Computer Science and Communication Engineering (ICCSCE 2025)
PB  - Atlantis Press
SP  - 603
EP  - 615
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-858-5_52
DO  - 10.2991/978-94-6463-858-5_52
ID  - Sameera2025
ER  -