adversarial examples
Adversarial examples refer to specially crafted inputs that are deliberately designed to mislead or deceive machine learning models. These input variations may appear indistinguishable from normal inputs to humans but can cause the model to produce incorrect or unexpected outputs.
Requires login.
Related Concepts (1)
Similar Concepts
- adversarial attacks
- adversarial deep learning
- adversarial detection and defense
- adversarial examples and attacks on ai systems
- adversarial feature learning
- adversarial image classification
- adversarial machine learning
- adversarial perturbations
- adversarial privacy attacks
- adversarial reinforcement learning
- adversarial risk analysis
- adversarial robustness
- adversarial text generation
- adversarial training
- adversary profiling