Improving Deep Learning Model Robustness Against Adversarial Attack by Increasing the Network Capacity

Marco Marchetti; Edmond S. L. Ho

doi:10.1007/978-3-031-21101-0_7

Improving Deep Learning Model Robustness Against Adversarial Attack by Increasing the Network Capacity

Marco Marchetti, Edmond S. L. Ho^*

^*Corresponding author for this work

Computer and Information Sciences Department

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

7 Downloads (Pure)

Abstract

Nowadays, we are more and more reliant on Deep Learning (DL) models and thus it is essential to safeguard the security of these systems. This paper explores the security issues in Deep Learning and analyses, through the use of experiments, the way forward to build more resilient models. Experiments are conducted to identify the strengths and weaknesses of a new approach to improve the robustness of DL models against adversarial attacks. The results show improvements and new ideas that can be used as recommendations for researchers and practitioners to create increasingly better DL algorithms.

Original language	English
Title of host publication	Advances in Cybersecurity, Cybercrimes, and Smart Emerging Technologies
Subtitle of host publication	CCSET 2022
Editors	Ahmed A. Abd El-Latif, Yassine Maleh, Wojciech Mazurczyk, Mohammed ELAffendi, Mohamed I. Alkanhal
Place of Publication	Cham, Switzerland
Publisher	Springer
Pages	85-96
Number of pages	12
ISBN (Electronic)	9783031211010
ISBN (Print)	9783031211003
DOIs	https://doi.org/10.1007/978-3-031-21101-0_7
Publication status	Published - 12 Mar 2023
Event	International conference on Cybersecurity, Cybercrimes, and Smart Emerging Technologies - Prince Sultan University, Saudi Arabia Duration: 10 May 2022 → 11 May 2022

Publication series

Name	Engineering Cyber-Physical Systems and Critical Infrastructures
Publisher	Springer
Volume	4
ISSN (Print)	2731-5002
ISSN (Electronic)	2731-5010

Conference

Conference	International conference on Cybersecurity, Cybercrimes, and Smart Emerging Technologies
Abbreviated title	CCSET2022
Country/Territory	Saudi Arabia
Period	10/05/22 → 11/05/22

Keywords

Machine Learning
Deep Learning
Security
Measurement
Perturbation methods
Robustness

Access to Document

10.1007/978-3-031-21101-0_7

AdversarialAccepted author manuscript, 135 KBLicence: CC BY

Cite this

Marchetti, M., & Ho, E. S. L. (2023). Improving Deep Learning Model Robustness Against Adversarial Attack by Increasing the Network Capacity. In A. A. A. El-Latif, Y. Maleh, W. Mazurczyk, M. ELAffendi, & M. I. Alkanhal (Eds.), Advances in Cybersecurity, Cybercrimes, and Smart Emerging Technologies: CCSET 2022 (pp. 85-96). (Engineering Cyber-Physical Systems and Critical Infrastructures ; Vol. 4). Springer. https://doi.org/10.1007/978-3-031-21101-0_7

Marchetti, Marco ; Ho, Edmond S. L. / Improving Deep Learning Model Robustness Against Adversarial Attack by Increasing the Network Capacity. Advances in Cybersecurity, Cybercrimes, and Smart Emerging Technologies: CCSET 2022. editor / Ahmed A. Abd El-Latif ; Yassine Maleh ; Wojciech Mazurczyk ; Mohammed ELAffendi ; Mohamed I. Alkanhal. Cham, Switzerland : Springer, 2023. pp. 85-96 (Engineering Cyber-Physical Systems and Critical Infrastructures ).

@inproceedings{fc2e8fe7ee574d72bd09978e51de92f4,

title = "Improving Deep Learning Model Robustness Against Adversarial Attack by Increasing the Network Capacity",

abstract = "Nowadays, we are more and more reliant on Deep Learning (DL) models and thus it is essential to safeguard the security of these systems. This paper explores the security issues in Deep Learning and analyses, through the use of experiments, the way forward to build more resilient models. Experiments are conducted to identify the strengths and weaknesses of a new approach to improve the robustness of DL models against adversarial attacks. The results show improvements and new ideas that can be used as recommendations for researchers and practitioners to create increasingly better DL algorithms.",

keywords = "Machine Learning, Deep Learning, Security, Measurement, Perturbation methods, Robustness",

author = "Marco Marchetti and Ho, {Edmond S. L.}",

year = "2023",

month = mar,

day = "12",

doi = "10.1007/978-3-031-21101-0_7",

language = "English",

isbn = "9783031211003",

series = "Engineering Cyber-Physical Systems and Critical Infrastructures ",

publisher = "Springer",

pages = "85--96",

editor = "{ El-Latif}, {Ahmed A. Abd} and Yassine Maleh and Wojciech Mazurczyk and Mohammed ELAffendi and Alkanhal, {Mohamed I.}",

booktitle = "Advances in Cybersecurity, Cybercrimes, and Smart Emerging Technologies",

address = "Germany",

note = "International conference on Cybersecurity, Cybercrimes, and Smart Emerging Technologies, CCSET2022 ; Conference date: 10-05-2022 Through 11-05-2022",

}

Marchetti, M & Ho, ESL 2023, Improving Deep Learning Model Robustness Against Adversarial Attack by Increasing the Network Capacity. in AAA El-Latif, Y Maleh, W Mazurczyk, M ELAffendi & MI Alkanhal (eds), Advances in Cybersecurity, Cybercrimes, and Smart Emerging Technologies: CCSET 2022. Engineering Cyber-Physical Systems and Critical Infrastructures , vol. 4, Springer, Cham, Switzerland, pp. 85-96, International conference on Cybersecurity, Cybercrimes, and Smart Emerging Technologies, Saudi Arabia, 10/05/22. https://doi.org/10.1007/978-3-031-21101-0_7

Improving Deep Learning Model Robustness Against Adversarial Attack by Increasing the Network Capacity. / Marchetti, Marco; Ho, Edmond S. L.
Advances in Cybersecurity, Cybercrimes, and Smart Emerging Technologies: CCSET 2022. ed. / Ahmed A. Abd El-Latif; Yassine Maleh; Wojciech Mazurczyk; Mohammed ELAffendi; Mohamed I. Alkanhal. Cham, Switzerland: Springer, 2023. p. 85-96 (Engineering Cyber-Physical Systems and Critical Infrastructures ; Vol. 4).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Improving Deep Learning Model Robustness Against Adversarial Attack by Increasing the Network Capacity

AU - Marchetti, Marco

AU - Ho, Edmond S. L.

PY - 2023/3/12

Y1 - 2023/3/12

N2 - Nowadays, we are more and more reliant on Deep Learning (DL) models and thus it is essential to safeguard the security of these systems. This paper explores the security issues in Deep Learning and analyses, through the use of experiments, the way forward to build more resilient models. Experiments are conducted to identify the strengths and weaknesses of a new approach to improve the robustness of DL models against adversarial attacks. The results show improvements and new ideas that can be used as recommendations for researchers and practitioners to create increasingly better DL algorithms.

AB - Nowadays, we are more and more reliant on Deep Learning (DL) models and thus it is essential to safeguard the security of these systems. This paper explores the security issues in Deep Learning and analyses, through the use of experiments, the way forward to build more resilient models. Experiments are conducted to identify the strengths and weaknesses of a new approach to improve the robustness of DL models against adversarial attacks. The results show improvements and new ideas that can be used as recommendations for researchers and practitioners to create increasingly better DL algorithms.

KW - Machine Learning

KW - Deep Learning

KW - Security

KW - Measurement

KW - Perturbation methods

KW - Robustness

U2 - 10.1007/978-3-031-21101-0_7

DO - 10.1007/978-3-031-21101-0_7

M3 - Conference contribution

SN - 9783031211003

T3 - Engineering Cyber-Physical Systems and Critical Infrastructures

SP - 85

EP - 96

BT - Advances in Cybersecurity, Cybercrimes, and Smart Emerging Technologies

A2 - El-Latif, Ahmed A. Abd

A2 - Maleh, Yassine

A2 - Mazurczyk, Wojciech

A2 - ELAffendi, Mohammed

A2 - Alkanhal, Mohamed I.

PB - Springer

CY - Cham, Switzerland

T2 - International conference on Cybersecurity, Cybercrimes, and Smart Emerging Technologies

Y2 - 10 May 2022 through 11 May 2022

ER -

Marchetti M, Ho ESL. Improving Deep Learning Model Robustness Against Adversarial Attack by Increasing the Network Capacity. In El-Latif AAA, Maleh Y, Mazurczyk W, ELAffendi M, Alkanhal MI, editors, Advances in Cybersecurity, Cybercrimes, and Smart Emerging Technologies: CCSET 2022. Cham, Switzerland: Springer. 2023. p. 85-96. (Engineering Cyber-Physical Systems and Critical Infrastructures ). Epub 2023 Mar 11. doi: 10.1007/978-3-031-21101-0_7

Improving Deep Learning Model Robustness Against Adversarial Attack by Increasing the Network Capacity

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this