Get in Touch
 Duration 14 hours

Course Outline

Foundations of Privacy-Preserving ML

  • Identifying risks and motivations in sensitive data contexts.
  • Survey of core privacy-preserving ML methodologies.
  • Threat modeling and regulatory frameworks (e.g., GDPR, HIPAA).

Federated Learning

  • Core concepts and system architecture.
  • Client-server synchronization and aggregation processes.
  • Practical implementation with PySyft and Flower.

Differential Privacy

  • Theoretical underpinnings and mathematical foundations.
  • Integration of DP into data querying and training loops.
  • Application using Opacus and TensorFlow Privacy.

Secure Multiparty Computation (SMPC)

  • Protocols and practical use cases for SMPC.
  • Comparing encryption-based and secret-sharing strategies.
  • Constructing secure workflows with CrypTen or PySyft.

Homomorphic Encryption

  • Distinguishing between fully and partially homomorphic schemes.
  • Executing encrypted inference for sensitive tasks.
  • Hands-on exercises with TenSEAL and Microsoft SEAL.

Industry Applications and Case Studies

  • Healthcare: Leveraging federated learning for medical AI.
  • Finance: Secure collaboration for risk modeling and compliance.
  • Defense and government sector implementations.

Requirements

  • Solid grasp of foundational machine learning principles.
  • Proficiency in Python and major ML frameworks (e.g., PyTorch, TensorFlow).
  • Basic knowledge of data privacy or cybersecurity standards is beneficial.

Target Audience

  • AI researchers and data scientists.
  • Professionals in data protection and privacy compliance.
  • Security engineers operating within regulated sectors.

Number of participants


Price per participant

Testimonials (3)

Upcoming Courses

Related Categories