Hybrid Attack

Hybrid attacks represent a growing concern in various machine learning and cybersecurity contexts, focusing on combining different attack strategies to exploit vulnerabilities more effectively. Current research investigates these attacks across diverse applications, including federated learning (using methods like hierarchical auditing and reinforcement learning to detect malicious clients), multi-exit neural networks (analyzing membership inference vulnerabilities and developing defenses), and bio-cybersecurity (detecting digitally encoded triggers in DNA sequences using deep learning). Understanding and mitigating these sophisticated attacks is crucial for enhancing the robustness and trustworthiness of machine learning systems and protecting sensitive data across multiple domains.

Papers

December 6, 2023

TrustFed: A Reliable Federated Learning Framework with Malicious-Attack Resistance
Hangn Su, Jianhong Zhou, Xianhua Niu, Gang Feng
Single Learning Multi Round Audit Model Hybrid Attack

November 15, 2023

On the Foundation of Distributionally Robust Reinforcement Learning
Shengbo Wang, Nian Si, Jose Blanchet, Zhengyuan Zhou
Optimal Policy Robust Policy Flawed Foundation Robust Reinforcement Learning Distributionally Robust Hybrid Attack

August 23, 2022

Auditing Membership Leakages of Multi-Exit Networks
Zheng Li, Yiyong Liu, Xinlei He, Ning Yu, Michael Backes, Yang Zhang
Membership Inference Attack Attack Performance Membership Privacy Multi Exit Hybrid Attack

February 23, 2022

Using Deep Learning to Detect Digitally Encoded DNA Trigger for Trojan Malware in Bio-Cyber Attacks
Mohd Siblee Islam, Stepan Ivanov, Hamdan Awan, Jennifer Drohan, Sasitharan Balasubramaniam, Lee Coffey, Srivatsan Kidambi, Witty Sri-saan
Deep Learning Cyber Attack Trojan Attack Protein Sequence Encoder Hybrid Attack

November 23, 2021

Adversarial machine learning for protecting against online manipulation
Stefano Cresci, Marinella Petrocchi, Angelo Spognardi, Stefano Tognazzi
Adversarial Example Adversarial Learning Image Recognition Social Bot Detection Market Manipulation Hybrid Attack

Hybrid Attack

Papers

TrustFed: A Reliable Federated Learning Framework with Malicious-Attack Resistance

On the Foundation of Distributionally Robust Reinforcement Learning

Auditing Membership Leakages of Multi-Exit Networks

Using Deep Learning to Detect Digitally Encoded DNA Trigger for Trojan Malware in Bio-Cyber Attacks

Adversarial machine learning for protecting against online manipulation