Breaking Machine Learning Models with Adversarial Attacks and its Variants

Machine learning models can be by adversarial attacks, subtle, imperceptible perturbations to inputs that cause the model to produce erroneous outputs. This tutorial introduces adversarial examples and its variants, explaining why even stateof-the-art models are vulnerable and how this impacts secu...

Full description

Saved in:
Bibliographic Details
Main Author: Pavan Reddy
Format: Article
Language:English
Published: LibraryPress@UF 2025-05-01
Series:Proceedings of the International Florida Artificial Intelligence Research Society Conference
Online Access:https://journals.flvc.org/FLAIRS/article/view/139042
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Machine learning models can be by adversarial attacks, subtle, imperceptible perturbations to inputs that cause the model to produce erroneous outputs. This tutorial introduces adversarial examples and its variants, explaining why even stateof-the-art models are vulnerable and how this impacts security in AI. It provides an overview of key concepts (such as black-box vs. white-box attack scenarios) and survey common attack techniques and defensive strategies. A hands-on component using Google Colab and the open-source Adversarial Lab toolkit allows attendees to craft adversarial examples and test model robustness in real time. Throughout, we emphasize both the practical skills and the ethical considerations needed to apply adversarial machine learning in a responsible manner. Attendees will gain a comprehensive foundation in adversarial attacks and insights into building more robust, secure machine learning models.
ISSN:2334-0754
2334-0762