ML Security

Model theft, adversarial inputs, data poisoning and ML supply chain.

6 concepts3 labs1 challenges

0% complete

Keep going!

Concepts

Start learning

Work through the concepts in order, or jump to what you need.

Concept · Hands-on lab included

Concept · Hands-on lab included

Concept · Hands-on lab included

Concept · Hands-on lab included

Concept · Hands-on lab included

Concept · Hands-on lab included

Practice

Apply what you've learned in guided, production-style environments.

40 min · intermediate · Adversarial Perturbations

35 min · advanced · Model Poisoning Attacks

40 min · intermediate · Model Theft and Extraction

Continue

Keep your learning streak going with related modules.

The full ML lifecycle and the operational discipline MLOps adds.

MLflow tracking, run metadata, reproducibility and comparing experiments.

DVC, data lineage and immutability for training data.