Adversarial Robustness for Machine Learning

Erscheinungsjahr 2022
ISBN: 978-0-12-824020-5
Verlag: William Andrew Publishing

Buch, Englisch, 298 Seiten, Format (B × H): 216 mm x 286 mm, Gewicht: 450 g

Adversarial Robustness for Machine Learning summarizes the recent progress on this topic and introduces popular algorithms on adversarial attack, defense and veri?cation. Sections cover adversarial attack, veri?cation and defense, mainly focusing on image classi?cation applications which are the standard benchmark considered in the adversarial robustness community. Other sections discuss adversarial examples beyond image classification, other threat models beyond testing time attack, and applications on adversarial robustness. For researchers, this book provides a thorough literature review that summarizes latest progress in the area, which can be a good reference for conducting future research.

In addition, the book can also be used as a textbook for graduate courses on adversarial robustness or trustworthy machine learning. While machine learning (ML) algorithms have achieved remarkable performance in many applications, recent studies have demonstrated their lack of robustness against adversarial disturbance. The lack of robustness brings security concerns in ML models for real applications such as self-driving cars, robotics controls and healthcare systems.

Zielgruppe

<p>Computer scientists and engineers at Universities; R&D engineers in industry.</p>

Autoren/Hrsg.

Hsieh, Cho-Jui

Fachgebiete

• Mathematik | Informatik EDV | Informatik Informatik Künstliche Intelligenz Maschinelles Lernen

Inhaltsverzeichnis

1. White-box attack
2. Soft-label Black-box Attack
3. Decision-based attack
4. Attack Transferibility
5. Attacks in the physical world
6. Convex relaxation Framework
7. Layer-wise relaxation (primal algorithms)
8. Dual approach
9. Probabilistic veri?cation
10. Adversarial training
11. Certi?ed defense
12. Randomization
13. Detection methods
14. Robustness of other machine learning models beyond neural networks
15. NLP models
16. Graph neural network
17. Recommender systems
18. Reinforcement Learning
19. Speech models
20. Multi-modal models
21. Backdoor attack and defense
22. Data poisoning attack and defense
23. Transfer learning
24. Explainability and interpretability
25. Representation learning
26. Privacy and watermarking

Über Autor(innen)

Hsieh, Cho-Jui
Dr. Cho-Jui Hsieh is an Assistant Professor at the UCLA Computer Science department. His research focuses on developing algorithms and optimization techniques for training large-scale and robust machine learning models. He publishes in top-tier machine learning conferences including ICML, NIPS, KDD, ICLR and has won the best paper awards at KDD 2010, ICDM 2012, ICPP 2018, best paper ?nalist at AISEC 2017 and best student paper ?nalist at SC 2019. He is also the author of several widely used open source machine learning software including LIBLINEAR. His work has been cited by more than 13,000 times on Google scholar.