August 26
15:00–15:30
Secure ML modeling
Track 2
Russian
The report talks about improving the security of machine learning models at various stages of development. It introduces a list of adversarial attacks and their use in model training and testing, as well as the security of the environment when training and running models.