From the course: MLOps Essentials: Model Deployment and Monitoring
Join today to access over 24,400 courses taught by industry experts.
Security of ML assets
From the course: MLOps Essentials: Model Deployment and Monitoring
Security of ML assets
- [Instructor] An important area in MLOps around responsible AI is the security of ML assets like data and models. Machine learning solutions also have threats of adversarial attacks that compromise the model as well as data used for training and inference. This is in addition to the other security threats that any other software application would face. Let's discuss the types of ML specific threats that such solutions face. There are two main assets for ML, data and model. When it comes to data, sharing across trust boundaries carries the risk of intrusion during transport, hackers can poison training data by introducing negative samples that influence model behavior, there is also the threat of data theft and break-ins which is becoming widespread these days, there is the risk of hackers reverse engineering redacted or deleted data to uncover private information. On the model side, there is a high risk of the…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.
