Privacy-preserving machine learning with TensorFlow
Who is this presentation for?
- Developers and data scientists with an intermediate or advanced understanding of machine learning
Today, we’re trying to take advantage of machine learning across many facets of modern life. However, many of our most impactful uses of machine learning in healthcare, transportation, and social life are blocked as they require access to sensitive data.
TF Encrypted and PySyft are complementary open source libraries for designing and building privacy-preserving machine learning workflows. They both extend TensorFlow and aim to make privacy-preserving machine learning easy without needing to understand the complexities of cryptography, distributed systems, or high-performance computing.
Jason Mancuso and Yann Dupis demonstrate how to build and deploy privacy-preserving machine learning models using TF Encrypted, PySyft-TensorFlow, and the TensorFlow ecosystem. You’ll learn how to use TF Encrypted and PySyft to train and deploy machine learning models using remote execution, secure federated learning, and encrypted predictions in the cloud while preserving the privacy of both the model and the end user’s input data. After an introduction to the landscape of privacy-preserving machine learning, Jason and Yann walk you through a series of hands-on exercises for building models with TF Encrypted’s secure primitives and PySyft-TensorFlow. Join in to gain the skills you need to identify use cases requiring heightened privacy and security and design, prototype, and deploy private machine learning.
- Experience with TensorFlow
- Familarity with Google Compute Engine (useful but not required)
Materials or downloads needed in advance
- Attendees will need to follow the instructions laid out at https://github.com/dropoutlabs/tf-world *BEFORE* they arrive onsite. This will primarily include some software installation, as well as setting up a Google Cloud account for those wishing to run examples in the cloud.
- At a minimum, we require attendees to have a usable Python environment (3.5+) with TensorFlow 2.x and recommend using a package manager like virtualenv or conda. For those interested in the cloud examples they must also have installed the Google Cloud SDK.
- Note that we will continue to update these instructions in the run-up to the conference and plan to provide automated installation scripts for most of these tasks. Attendees should check back in the days leading up to the tutorial to finalize their environments.
What you'll learn
- Understand the lifecycle of a machine learning model, where and how this lifecycle leaks data and model privacy, and how TF Encrypted can be used to prevent these leaks
Jason Mancuso is a research scientist at Dropout Labs, the founder of Cleveland AI, and an active member of the AI Village at DEF CON and the OpenMined community. He works on novel methods of making machine learning more performant for privacy-preserving techniques, most notably by contributing to the TF Encrypted project. He’s worked on a variety of safety and security problems, including safe reinforcement learning, secure and verifiable agent auditing, and neural network robustness. His work with the Cleveland Clinic established a state-of-the-art blood test classification and demonstrated that machine learning can virtually eliminate the problem of medical malpractice due to contaminated blood samples.
Yann Dupis is a machine learning engineer and privacy researcher at Dropout Labs. Previously, he was an actuary at the largest insurance company in Canada in reinsurance and then in research and development, and he managed a data science team at Deloitte in San Francisco, working with several Fortune 500 enterprises in the consumer and product industry. He holds an MASc in electrical and computer engineering from Institut Superieur d’Electronique de Paris. In his free time, you can find him surfing at Ocean Beach or indoor rock climbing in San Francisco.
For conference registration information and customer service
For more information on community discounts and trade opportunities with O’Reilly conferences
For information on exhibiting or sponsoring a conference
For media/analyst press inquires