Autoencoder-based feature learning for cyber security applications

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contributionResearchpeer-review

Abstract

This paper presents a novel feature learning model for cyber security tasks. We propose to use Auto-encoders (AEs), as a generative model, to learn latent representation of different feature sets. We show how well the AE is capable of automatically learning a reasonable notion of semantic similarity among input features. Specifically, the AE accepts a feature vector, obtained from cyber security phenomena, and extracts a code vector that captures the semantic similarity between the feature vectors. This similarity is embedded in an abstract latent representation. Because the AE is trained in an unsupervised fashion, the main part of this success comes from appropriate original feature set that is used in this paper. It can also provide more discriminative features in contrast to other feature engineering approaches. Furthermore, the scheme can reduce the dimensionality of the features thereby signicantly minimising the memory requirements. We selected two different cyber security tasks: networkbased anomaly intrusion detection and Malware classication. We have analysed the proposed scheme with various classifiers using publicly available datasets for network anomaly intrusion detection and malware classifications. Several appropriate evaluation metrics show improvement compared to prior results.

LanguageEnglish
Title of host publicationInternational Joint Conference on Neural Networks
Place of PublicationPiscataway, NJ
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Pages3854-3861
Number of pages8
ISBN (Electronic)9781509061822
ISBN (Print)9781509061815, 9781509061839
DOIs
Publication statusPublished - 30 Jun 2017
EventInternational Joint Conference on Neural Networks 2017 - Anchorage, AK, United States
Duration: 14 May 201719 May 2017
Conference number: 17010725

Conference

ConferenceInternational Joint Conference on Neural Networks 2017
Abbreviated titleIJCNN
CountryUnited States
CityAK
Period14/05/1719/05/17

Fingerprint

Intrusion detection
Semantics
Classifiers
Data storage equipment
Malware

Cite this

Yousefiazar, M., Varadharajan, V., Hamey, L., & Tupakula, U. (2017). Autoencoder-based feature learning for cyber security applications. In International Joint Conference on Neural Networks (pp. 3854-3861). [7966342] Piscataway, NJ: Institute of Electrical and Electronics Engineers (IEEE). https://doi.org/10.1109/IJCNN.2017.7966342
Yousefiazar, Mahmood ; Varadharajan, Vijayaraghavan ; Hamey, Leonard ; Tupakula, Udaya. / Autoencoder-based feature learning for cyber security applications. International Joint Conference on Neural Networks. Piscataway, NJ : Institute of Electrical and Electronics Engineers (IEEE), 2017. pp. 3854-3861
@inproceedings{1993416244f040bb82163428ae7209f6,
title = "Autoencoder-based feature learning for cyber security applications",
abstract = "This paper presents a novel feature learning model for cyber security tasks. We propose to use Auto-encoders (AEs), as a generative model, to learn latent representation of different feature sets. We show how well the AE is capable of automatically learning a reasonable notion of semantic similarity among input features. Specifically, the AE accepts a feature vector, obtained from cyber security phenomena, and extracts a code vector that captures the semantic similarity between the feature vectors. This similarity is embedded in an abstract latent representation. Because the AE is trained in an unsupervised fashion, the main part of this success comes from appropriate original feature set that is used in this paper. It can also provide more discriminative features in contrast to other feature engineering approaches. Furthermore, the scheme can reduce the dimensionality of the features thereby signicantly minimising the memory requirements. We selected two different cyber security tasks: networkbased anomaly intrusion detection and Malware classication. We have analysed the proposed scheme with various classifiers using publicly available datasets for network anomaly intrusion detection and malware classifications. Several appropriate evaluation metrics show improvement compared to prior results.",
author = "Mahmood Yousefiazar and Vijayaraghavan Varadharajan and Leonard Hamey and Udaya Tupakula",
year = "2017",
month = "6",
day = "30",
doi = "10.1109/IJCNN.2017.7966342",
language = "English",
isbn = "9781509061815",
pages = "3854--3861",
booktitle = "International Joint Conference on Neural Networks",
publisher = "Institute of Electrical and Electronics Engineers (IEEE)",
address = "United States",

}

Yousefiazar, M, Varadharajan, V, Hamey, L & Tupakula, U 2017, Autoencoder-based feature learning for cyber security applications. in International Joint Conference on Neural Networks., 7966342, Institute of Electrical and Electronics Engineers (IEEE), Piscataway, NJ, pp. 3854-3861, International Joint Conference on Neural Networks 2017, AK, United States, 14/05/17. https://doi.org/10.1109/IJCNN.2017.7966342

Autoencoder-based feature learning for cyber security applications. / Yousefiazar, Mahmood; Varadharajan, Vijayaraghavan; Hamey, Leonard; Tupakula, Udaya.

International Joint Conference on Neural Networks. Piscataway, NJ : Institute of Electrical and Electronics Engineers (IEEE), 2017. p. 3854-3861 7966342.

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contributionResearchpeer-review

TY - GEN

T1 - Autoencoder-based feature learning for cyber security applications

AU - Yousefiazar, Mahmood

AU - Varadharajan, Vijayaraghavan

AU - Hamey, Leonard

AU - Tupakula, Udaya

PY - 2017/6/30

Y1 - 2017/6/30

N2 - This paper presents a novel feature learning model for cyber security tasks. We propose to use Auto-encoders (AEs), as a generative model, to learn latent representation of different feature sets. We show how well the AE is capable of automatically learning a reasonable notion of semantic similarity among input features. Specifically, the AE accepts a feature vector, obtained from cyber security phenomena, and extracts a code vector that captures the semantic similarity between the feature vectors. This similarity is embedded in an abstract latent representation. Because the AE is trained in an unsupervised fashion, the main part of this success comes from appropriate original feature set that is used in this paper. It can also provide more discriminative features in contrast to other feature engineering approaches. Furthermore, the scheme can reduce the dimensionality of the features thereby signicantly minimising the memory requirements. We selected two different cyber security tasks: networkbased anomaly intrusion detection and Malware classication. We have analysed the proposed scheme with various classifiers using publicly available datasets for network anomaly intrusion detection and malware classifications. Several appropriate evaluation metrics show improvement compared to prior results.

AB - This paper presents a novel feature learning model for cyber security tasks. We propose to use Auto-encoders (AEs), as a generative model, to learn latent representation of different feature sets. We show how well the AE is capable of automatically learning a reasonable notion of semantic similarity among input features. Specifically, the AE accepts a feature vector, obtained from cyber security phenomena, and extracts a code vector that captures the semantic similarity between the feature vectors. This similarity is embedded in an abstract latent representation. Because the AE is trained in an unsupervised fashion, the main part of this success comes from appropriate original feature set that is used in this paper. It can also provide more discriminative features in contrast to other feature engineering approaches. Furthermore, the scheme can reduce the dimensionality of the features thereby signicantly minimising the memory requirements. We selected two different cyber security tasks: networkbased anomaly intrusion detection and Malware classication. We have analysed the proposed scheme with various classifiers using publicly available datasets for network anomaly intrusion detection and malware classifications. Several appropriate evaluation metrics show improvement compared to prior results.

UR - http://www.scopus.com/inward/record.url?scp=85030984896&partnerID=8YFLogxK

U2 - 10.1109/IJCNN.2017.7966342

DO - 10.1109/IJCNN.2017.7966342

M3 - Conference proceeding contribution

SN - 9781509061815

SN - 9781509061839

SP - 3854

EP - 3861

BT - International Joint Conference on Neural Networks

PB - Institute of Electrical and Electronics Engineers (IEEE)

CY - Piscataway, NJ

ER -

Yousefiazar M, Varadharajan V, Hamey L, Tupakula U. Autoencoder-based feature learning for cyber security applications. In International Joint Conference on Neural Networks. Piscataway, NJ: Institute of Electrical and Electronics Engineers (IEEE). 2017. p. 3854-3861. 7966342 https://doi.org/10.1109/IJCNN.2017.7966342