Please join us for a UW Data Science Seminar event on Tuesday, November 8th from 4:30 to 5:20 p.m. PDT. The seminar will feature Ian Quah, a software engineer at Duality Technologies.
Use this zoom link to join
“Homomorphic Encryption for Encrypted Machine Learning through the OpenFHE library”
Abstract: The question of privacy in the age of big data is the subject of great debate. The success of modern deep learning can be attributed to increased data availability, but in the context of sensitive data, this comes at a cost to individual data owners. With regulations such as GDPR, PIPL, and CCPA, it is clear that governments are becoming increasingly concerned with how their citizens’ data are used. However, what if it was possible to train machine learning models without seeing the underlying data? Homomorphic Encryption, often called the “holy grail” of cryptography, allows us to do just that. Through Homomorphic Encryption, we can carry out mathematical operations on encrypted data as if we were working on plaintext data. This ability to compute on encrypted data enables us to train encrypted models on encrypted data, providing strong privacy guarantees.
My goal for this talk is to introduce the data science community to fully homomorphic Encryption and the OpenFHE library. I do this by first introducing core concepts that a data science practitioner should know before walking through an example of the OpenFHE library in action.
Biography: Ian Quah is a software engineer at Duality Technologies, working on tasks ranging from cryptography to encrypted machine learning. Additionally, Ian is a maintainer for the OpenFHE library, where his work focuses on technical outreach, documentation, and community growth. Ian’s research interests include privacy-preserving machine learning, deep reinforcement learning, and biologically-plausible deep learning.
The UW Data Science Seminar is an annual lecture series at the University of Washington that hosts scholars working across applied areas of data science, such as the sciences, engineering, humanities and arts along with methodological areas in data science, such as computer science, applied math and statistics. Our presenters come from all domain fields and include occasional external speakers from regional partners, governmental agencies and industry.
The 2022-2023 seminars will be virtual, and are free and open to the public.