Optimal Adversarial

Optimal adversarial research focuses on finding the most effective strategies for attacking and defending machine learning models, particularly deep neural networks, against adversarial examples—inputs designed to cause misclassification. Current research explores this through various frameworks, including multi-agent reinforcement learning, optimal transport theory, and evolutionary algorithms, often employing techniques like adversarial training and game theory to find optimal attack and defense strategies under different cost constraints. This work is crucial for improving the robustness and security of machine learning systems across diverse applications, from transportation networks to image classification, by identifying vulnerabilities and developing more resilient models.

Papers

January 17, 2024

An Optimal Transport Approach for Computing Adversarial Training Lower Bounds in Multiclass Classification
Nicolas Garcia Trillos, Matt Jacobs, Jakwang Kim, Matthew Werenski
Native Robustness Adversarial Training Optimal Transport Multiclass Classification Optimal Adversarial Multi Marginal Optimal Transport

December 22, 2023

Multi-Agent Reinforcement Learning for Assessing False-Data Injection Attacks on Transportation Networks
Taha Eghtesad, Sirui Li, Yevgeniy Vorobeychik, Aron Laszka
Multi Agent Reinforcement Learning Road Network Transportation Network Optimal Adversarial Injection Attack

November 1, 2023

Optimal Cost Constrained Adversarial Attacks For Multiple Agent Systems
Ziqing Lu, Guanlin Liu, Lifeng Lai, Weiyu Xu
Adversarial Attack Multi Agent Multi Agent System Malicious Agent Optimal Adversarial Optimal Cost

October 31, 2023

Best of Both Worlds Guarantees for Smoothed Online Quadratic Optimization
Neelkamal Bhuyan, Debankur Mukherjee, Adam Wierman
Online Optimization Optimal Adversarial Asymptotic Performance Global Guarantee

July 13, 2023

Multi-objective Evolutionary Search of Variable-length Composite Semantic Perturbations
Jialiang Sun, Wen Yao, Tingsong Jiang, Xiaoqian Chen
Adversarial Attack Evolutionary Multi Objective Optimization Semantic Adversarial Optimal Adversarial Semantic Perturbation

June 13, 2023

On Achieving Optimal Adversarial Test Error
Justin D. Li, Matus Telgarsky
Adversarial Loss Adversarial Testing Optimal Adversarial

February 27, 2023

Implicit Poisoning Attacks in Two-Agent Reinforcement Learning: Adversarial Policies for Training-Time Attacks
Mohammad Mohammadi, Jonathan Nöther, Debmalya Mandal, Adish Singla, Goran Radanovic
Poisoning Attack Training Time Attack Adversarial Policy Optimal Adversarial Optimal Attack

December 10, 2022

Targeted Adversarial Attacks on Deep Reinforcement Learning Policies via Model Checking
Dennis Gross, Thiago D. Simao, Nils Jansen, Guillermo A. Perez
Adversarial Attack Deep Reinforcement Learning Adversarial Noise Model Checking Optimal Adversarial

August 15, 2022

A Multi-objective Memetic Algorithm for Auto Adversarial Attack Optimization Design
Jialiang Sun, Wen Yao, Tingsong Jiang, Xiaoqian Chen
Adversarial Attack Memetic Algorithm Optimal Adversarial Efficient Adversarial Attack

July 17, 2022

Achieve Optimal Adversarial Accuracy for Adversarial Deep Learning using Stackelberg Game
Xiao-Shan Gao, Shuang Liu, Lijia Yu
Adversarial Attack Adversarial Learning Game Theory Stackelberg Game Deep Equilibrium Optimal Adversarial

June 23, 2022

A Framework for Understanding Model Extraction Attack and Defense
Xun Xian, Mingyi Hong, Jie Ding
New Framework Adversarial Learning Model Extraction Attack Model Improvement Machine Learning a a Service Optimal Adversarial

May 16, 2022

Attacking and Defending Deep Reinforcement Learning Policies
Chao Wang
Adversarial Attack Deep Reinforcement Learning Robust Optimization Defense Algorithm Optimal Adversarial

April 27, 2022

The Multimarginal Optimal Transport Formulation of Adversarial Multiclass Classification
Nicolas Garcia Trillos, Matt Jacobs, Jakwang Kim
Generative Adversarial Adversarial Risk Learning Problem Multi Marginal Optimal Transport Optimal Adversarial Barycenter Driven Localization

November 24, 2021

Challenges of Adversarial Image Augmentations
Arno Blaas, Xavier Suau, Jason Ramapuram, Nicholas Apostoloff, Luca Zappella
Technical Challenge Adversarial AutoEncoder Adversarial Augmentation Augmentation Policy Optimal Adversarial Random Augmentation Random Data Augmentation

Optimal Adversarial

Papers

An Optimal Transport Approach for Computing Adversarial Training Lower Bounds in Multiclass Classification

Multi-Agent Reinforcement Learning for Assessing False-Data Injection Attacks on Transportation Networks

Optimal Cost Constrained Adversarial Attacks For Multiple Agent Systems

Best of Both Worlds Guarantees for Smoothed Online Quadratic Optimization

Multi-objective Evolutionary Search of Variable-length Composite Semantic Perturbations

On Achieving Optimal Adversarial Test Error

Implicit Poisoning Attacks in Two-Agent Reinforcement Learning: Adversarial Policies for Training-Time Attacks

Targeted Adversarial Attacks on Deep Reinforcement Learning Policies via Model Checking

A Multi-objective Memetic Algorithm for Auto Adversarial Attack Optimization Design

Achieve Optimal Adversarial Accuracy for Adversarial Deep Learning using Stackelberg Game

A Framework for Understanding Model Extraction Attack and Defense

Attacking and Defending Deep Reinforcement Learning Policies

The Multimarginal Optimal Transport Formulation of Adversarial Multiclass Classification

Challenges of Adversarial Image Augmentations