Trusted-AI/adversarial-robustness-toolbox

Support for additional attack model types in inference attacks

abigailgold opened this issue · 1 comments

Blackbox membership inference attack currently supports 3 model types: nn, rf and gb.
Blackbox attribute inference attack currently supports 2 model types: nn and rf.
Both could benefit from supporting additional attack model types (e.g., KNN, LR, etc.)

Also need to replace sklearn MLP implementation in attribute inference blackbox to pytorch (to improve performance).