Resources
Below is a directory of resources. The directory is not comprehensive and inclusion is not an endorsement
from NIST.
If you would like to suggest a resource to be added, please contact NIST scientist
Gary Howarth.
Quick Links (this page):
Open Source Attack Libraries:
(Adversarial Robustness Toolbox)- Membership Inference (Black Box) (DT, NN)
- Attribute Inference (Black Box) (DT, NN)
- Attribute Inference (white box) (scikit DT)
- Model Inversion (White Box) (NN)
- Row Reconstruction (White Box) (DT, NN)
- Other Attacks
- Supported Estimators Not all estimators work with all attacks.
- Membership Inference (Black Box) (NN) [Metrics Module]
- Supported Estimators These can be extended easily to support other estimators.
- Reference MI attacks
- Membership Inference (Black Box) (NN) [(All membership attacks)
- Secret Sharer Attack (White Box) (NN)
Research Codebases:
EvaluatingDPML- Membership Inference (Black Box) (NN)
- Attribute InferenceSecret Sharer Attack (Black Box And White Box) (NN)
- Paper
- Membership Inference (Black Box) (NN)
- Evasion and Poisoning attacks only
- Membership inference in federated learning settings
- Data reconstruction using gradient
- Examples
- Data reconstruction using gradients
Other Resources:
- Membership Inference Attacks and Defenses on Machine Learning Models Literature
- Membership Inference Competition MICO Microsoft
- Awesome ML Privacy Attacks (has code links)
- Survey ML Attacks paper
- Membership attacks on tradition models
- Feature Inference Attack on Model Predictions in Vertical Federated Learning,
- Privacy Preserving Machine Learning Resources
- Overview of Fed Deep Learning Privacy Attacks and Defensive Strategies
- PrivacyFL: A simulator for privacy-preserving and secure federated learning
- Client Privacy Leakage attacks in FL
- Fairness and Privacy-Preserving in Federated Learning: A Survey
- Inverting Gradient in FL
- Unleashing the Tiger: Inference Attacks on Split Learning
- From Gradient Leakage to Adversarial Attacks in Federated Learning
- Presentation: Privacy attacks on Decision Trees
- Membership Inference Attacks on Federated Horizontal Gradient Boosted Decision Trees
- https://dl.acm.org/doi/full/10.1145/3624010
Membership Inference:
- Membership Inference Attacks Against Machine Learning Models
- Evaluating Membership Inference Attacks and Defenses in Federated Learning