Now more than ever, the trade-off between privacy and the benefits of optimally training your algorithms is under heavy discussion. Luckily, companies (mainly Google), heavily invest in research in this field which resulted in open sourcing various libraries. In this talk we will go into the technical details or differential privacy, federated learning in Tensorflow and Multi-Party Compute.
The digital advertising industry has been a great driving factor in the recent developments in data science and AI. However, now more than ever, the trade-off between privacy and the benefits of optimally training your algorithms is under heavy discussion. Luckily, companies (mainly Google), heavily invest in research in this field which resulted in open sourcing various libraries.
After a brief introduction about the context privacy in online advertising, we'll dive into the technical details of:
-Differential privacy -Federated learning in Tensorflow -Multi-Party Compute -Chromium's Privacy Sandbox -My own vision around Private and Public layers present in a blog from 2018