Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List

Full text release has been delayed at the author's request until July 31, 2026

ETD Abstract Container

Abstract Header

Class-Based Adversarial Training for AI Robustness

Jost, Deirdre
http://orcid.org/0009-0002-7670-1164
http://rave.ohiolink.edu/etdc/view?acc_num=miami1722032769467779

Abstract Details

2024, Master of Science in Computer Science, Miami University, Computer Science and Software Engineering.
Adversarial training (AT) is a defense technique used to increase the robustness of neural networks. AT generates adversarial examples that maximize the loss to the model and then adjusts model parameters to minimize that loss. Previous AT methods typically use only a single attack to perturb adversarial examples that maximize loss, and ignore the roles that different image-classes play in determining final robustness. These techniques are thus unable to properly explore the perturbation space and cannot target specific weaknesses of the training data. As a result, they train models with diminished robustness. This thesis proposes class-based adversarial training, which increases the robustness of AT by using a variety of attacks that target the weakest image-classes of the dataset. We designed and implemented two novel algorithms within this category: the Various Attacks (VA) technique and the Advanced Adversarial Distributional Training (ADT++) technique. Using a novel testing framework created to better examine model robustness across a variety of metrics, we conducted a series of experiments on two benchmark datasets. The results demonstrate the superiority of the VA and ADT++ frameworks over state-of-the-art adversarial training methods.
Samer Khamaiseh (Advisor)
Honglu Jiang (Committee Member)
Hakam Alomari (Committee Member)
79 p.
Computer Science
machine learning; artificial intelligence; ai security; neural networks

Recommended Citations

Citations

  • Jost, D. (2024). Class-Based Adversarial Training for AI Robustness [Master's thesis, Miami University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=miami1722032769467779

    APA Style (7th edition)

  • Jost, Deirdre. Class-Based Adversarial Training for AI Robustness. 2024. Miami University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=miami1722032769467779.

    MLA Style (8th edition)

  • Jost, Deirdre. "Class-Based Adversarial Training for AI Robustness." Master's thesis, Miami University, 2024. http://rave.ohiolink.edu/etdc/view?acc_num=miami1722032769467779

    Chicago Manual of Style (17th edition)

miami1722032769467779
© 2024, some rights reserved.Creative Commons License Class-Based Adversarial Training for AI Robustness by Deirdre Jost is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. Based on a work at etd.ohiolink.edu.
This open access ETD is published by Miami University and OhioLINK.