- ホーム
- > 洋書
- > 英文書
- > Computer / General
Full Description
This book introduces a unified two-dimensional framework approach to adversarial machine learning—a central pillar to defending deep learning and large language model (LLM) systems against backdoor attacks, adversarial examples, and other deliberate manipulation.Spanning the machine learning lifecycle (pre-training to inference) and three major attack paradigms (backdoor, weight, and adversarial example), the framework introduced is built on the core principles of stealthiness, benign consistency, and adversarial inconsistency. Written for graduate students, researchers, and practitioners, the book also covers constructive applications (copyright and privacy protection, data-abuse mitigation, LLM alignment) and the challenges of generative AI and autonomous agents.
-
- 電子書籍
- ラズパイPico W 本格入門 wit…
-
- 電子書籍
- コミックヘヴン 2026年 11/10…
-
- 電子書籍
- 初めてのひよこクラブ2026年秋号
-
- 電子書籍
- フラジャイル・モーメント 気候危機の真…



