Theses and Dissertations
Date of Award
12-2020
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Computer Science
First Advisor
Dr. Hansheng Lei
Second Advisor
Dr. Fitratullah Khan
Third Advisor
Dr. Mahmoud K. Quweider
Abstract
In this thesis, a targeted adversarial attack is explored on a Support Vector Machine (SVM). SVM is defined by creating a separating boundary between two classes. Using a target class, any input can be modified to cross the “boundary line,” making the model predict the target class. To limit the modification, a percentage of an image of the target class is used to get several random sections. Using these sections, the input will be moved in small steps closer to the boundary point. The section that took the least number of steps to cause the model to predict the target class will be considered the optimal section. This method of attack can lead us to find the predominant features that the SVM uses to classify the target class. This knowledge can be used for further attacks and to find further vulnerabilities of the SVM model.
Recommended Citation
Rodriguez, Yessenia, "A Targeted Adversarial Attack on Support Vector Machine Using the Boundary Line" (2020). Theses and Dissertations. 756.
https://scholarworks.utrgv.edu/etd/756
Comments
Copyright 2020 Yessenia Rodriguez. All Rights Reserved.
https://go.openathens.net/redirector/utrgv.edu?url=https://www.proquest.com/dissertations-theses/targeted-adversarial-attack-on-support-vector/docview/2560025280/se-2?accountid=7119