Introduction to Privacy Preserving Machine Learning

Overview

This one-hour live webinar will introduce participants to the fundamentals of Privacy Preserving Machine Learning (PPML). The session will introduce key PPML concepts such as Federated Learning, Differential Privacy, and Homomorphic Encryption, giving participants a foundational understanding of how to balance privacy and transparency with the effectiveness of ML models. During the webinar, attendees will get practical insights into integrating privacy-preserving techniques into ML workflows using PySyft - an open source tool for secure and private machine learning.

Register here

Objectives

1. Understand the core concepts and the importance of PPML.
2. Learn the basics of Federated Learning, Differential Privacy, and Homomorphic Encryption.
3. Learn how PySyft enables privacy-preserving Machine learning

Audience

This webinar is designed for data scientists, AI practitioners, machine learning engineers, and developers interested in applying privacy-preserving techniques to their ML models, especially those working with sensitive data in domains like healthcare, finance, and government.

When & When

• Date: Thursday, 14 November 2024
• Time: 5 PM GMT / 6 PM CET / 12 PM EST / 9 AM PST
• Duration: 1 hour
• Location: Online (Information will be shared with attendees after registration)

Agenda

1. Introduction to PPML and PySyft (10 minutes):
   • Importance of privacy in Machine learning.
   • Overview of PySyft as a tool for privacy-preserving ML development.
2. Core PPML Methods (30 minutes):
   • Federated Learning:
     Training models across decentralized data sources.
   • Differential Privacy:
     Adding noise to data to maintain individual privacy in ML models.
   • Homomorphic Encryption:
     Secure computations on encrypted data.
3. PPML in PySyft with Structured Transparency (10 minutes):
   • PPML in practice using PySyft, and the Structured Transparency framework
4. Q&A Session (10 minutes):
   • Addressing participant questions and insights.

Takeaways

Participants will leave with a solid understanding of PPML, its importance, and how it can be applied to machine learning or data science workflows.

This webinar is an ideal starting point for professionals seeking hands-on tools like PySyft to ensure data privacy while leveraging the full potential of machine learning in sensitive environments.