EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Machine Learning and Deep Learning approaches in Network Intrusion Detection

Anil K. Makhija

Faculty Publications · 2021

DOI: 10.62458/021024-02 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Faculty Publications, 2021. Verfasst von Anil K. Makhija. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper presents multiple hybrid models for Intrusion detection systems (IDS). Some of the proposed models use combination of information-gain based feature selection followed by classification using Random Forests and Naïve Bayes algorithms. Some of the proposed model use combination of expectation-maximization based clustering, information-gain based feature selection and then feed forward neural network with the backpropagation training algorithm. NSL-KDD dataset has been used to train and validate the model and NSL-KDD Test dataset is used to test the accuracy, precision, recall and F1-score of each of the proposed model. Performance of the proposed models is also compared with performance of Random Forests and Naïve Bayes based classification. The experimental results on the model that uses combination of expectation-maximization based clustering, information-gain based feature selection and then feed forward neural network showed promising results on detecting the intrusion when tested on NSL-KDD Test dataset.

Zitieren

APA

Makhija, A. K. (2021). Machine Learning and Deep Learning approaches in Network Intrusion Detection. Faculty Publications. https://doi.org/10.62458/021024-02

Vancouver

Makhija AK. Machine Learning and Deep Learning approaches in Network Intrusion Detection. Faculty Publications. 2021. doi:10.62458/021024-02

BibTeX
@article{Makhija2021,
  author = {Makhija, Anil K.},
  title = {Machine Learning and Deep Learning approaches in Network Intrusion Detection},
  journal = {Faculty Publications},
  year = {2021},
  doi = {10.62458/021024-02}
}
Kostenlos starten →