Privacy-Preserving Machine Learning - Sivakanth Gopi, Microsoft Research
HTML-код
- Опубликовано: 3 июл 2024
- Slides: speakerdeck.com/aakashnsjovia...
Machine learning models are usually trained on sensitive user data, which can lead to privacy violations. In this talk, I will talk about 'Differential Privacy' (DP), a technique to learn broad statistics and train ML models on sensitive user data without compromising privacy. DP is increasingly becoming the gold standard for privacy, with even the US government using DP for releasing its latest census data.
Topics Covered
00:00 - Introduction & Motivation
04:23 - Differential Privacy
08:10 - Recipe for DP Algorithms
14:24 - Differentially Private ML
17:27 - Differentially Private SGD
20:30 - Results & Conclusion
Speaker: Sivakanth Gopi, Senior Researcher, Algorithms, Microsoft Research
Join our WhatsApp community: jovian.com/whatsapp
can anyone help me with this code ?
Hello, can you share the code in Discussions in the relevant course page?