Model Obfuscation

ModelObfuscator: Obfuscating Model Information to Protect Deployed ML-Based Systems
A novel approach that can significantly increase the difficulty of understanding the model information for human and automatic tools. It will not affect the model performance and only introduce small overheads.
Mingyi Zhou, Xiang Gao, Jing Wu, John Grundy, Xiao Chen, Chunyang Chen, Li Li
PDF Cite Code DOI
ModelObfuscator: Obfuscating Model Information to Protect Deployed ML-Based Systems