Security Testing: Adversarial Attacks, Poisoning & Model Inversion

Master adversarial machine learning security testing, data poisoning defense, and privacy-preserving mitigations under MITRE ATLAS and ISO/IEC TR 24028. Covers evasion attacks (FGSM, PGD), training corpus poisoning, membership inference, model inversion, extraction attacks, and differential privacy controls.