Network Anomaly Detection System Using Hidden Layer Information of Autoencoder 


Vol. 47,  No. 9, pp. 1310-1321, Sep.  2022
10.7840/kics.2022.47.9.1310


PDF
  Abstract

As Internet usage has been increasing, the importance of network intrusion detection systems (NIDS) has been highlighted. A promising solution for the NIDS is an autoencoder, a type of deep learning model. The conventional autoencoder uses only the input and output layers to detect intrusion, which draws a limitation. In this case, the information embedded in hidden layers would be ignored. The hidden layers of the autoencoder should be included in the detection process since they have information about the data. In order to overcome such limitations, we propose a novel anomaly detection solution that utilizes not only the input and output layers of the autoencoder but also hidden layers of the autoencoder to improve the detection performance. To evaluate the detection performance of the proposed solution, we use two popular network intrusion data sets and compare our solution with existing state-of-the-art methods. As a result, we confirm that the proposed solution outperforms other comparison methods. Specifically, our solution shows as high as 98% Accuracy and F1-score on average, while the comparison method shows 80% Accuracy and F1-score on average.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Related Articles
  Cite this article

[IEEE Style]

M. Kim, H. Kye, M. Kwon, "Network Anomaly Detection System Using Hidden Layer Information of Autoencoder," The Journal of Korean Institute of Communications and Information Sciences, vol. 47, no. 9, pp. 1310-1321, 2022. DOI: 10.7840/kics.2022.47.9.1310.

[ACM Style]

Miru Kim, Hyoseon Kye, and Minhae Kwon. 2022. Network Anomaly Detection System Using Hidden Layer Information of Autoencoder. The Journal of Korean Institute of Communications and Information Sciences, 47, 9, (2022), 1310-1321. DOI: 10.7840/kics.2022.47.9.1310.

[KICS Style]

Miru Kim, Hyoseon Kye, Minhae Kwon, "Network Anomaly Detection System Using Hidden Layer Information of Autoencoder," The Journal of Korean Institute of Communications and Information Sciences, vol. 47, no. 9, pp. 1310-1321, 9. 2022. (https://doi.org/10.7840/kics.2022.47.9.1310)

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880