Introduction to Responsible AI: Implement Ethical AI Using Python

Index

A

Achieving reliability, challenges

data quality

model drift

uncertainty

Achieving robustness, challenges

adversarial examples, transferability

model overfitting

outliers and noise

sensitivity, input variations

Adversarial attacks

Adversarial Debiasing algorithm

Adversarial examples

addressing ways

adversarial training

input preprocessing

model diversity

randomized defenses

“black-box” attacks

implications

defense challenges

model ensemble vulnerability

wider attack surface

transferability

Adversarial robustness

Adversarial testing

Adversarial training

AI-driven algorithms

AI Fairness 360 toolkit

AI security risks mitigation

backdoor detection and removal

conclusion

defense mechanisms, adversarial training

feature squeezing

gradient ...