Securing Classifiers Against Both White-Box and Black-Box Attacks using Encrypted-Input Obfuscation.
Giovanni Di CrescenzoBrian A. CoanLisa BahlerKurt RohloffYuriy PolyakovDavid Bruce CousinsPublished in: CCSW@CCS (2020)
Keyphrases
- black box
- white box
- black boxes
- source code
- white box testing
- training data
- support vector
- decision trees
- hybrid systems
- integration testing
- test cases
- reverse engineering
- source code metrics
- computer security
- rule extraction
- meta learning
- machine learning algorithms
- encryption scheme
- state transition
- training samples
- feature set
- training set
- high level