Anastasia Pustozerova @Vienna Deep Learning Meetup
Anastasia Pustozerova is researcher at SBA Research and gave an interesting talk on Differential Privacy for Machine Learning.
Talk Abstract
Machine Learning requires a lot of data to train effective models. Data owners might not be willing to share the data because of its private nature. Differential Privacy can help to quantify and enhance the privacy of sensitive data, when performing data analysis or machine learning. Companies including Apple, Google, Microsoft and even governments employ Differential Privacy to gain insights into the data while preserving the privacy of individual users. But what is Differential Privacy? How can it make machine learning more private? Most importantly, does it actually always guarantee the privacy of individuals? This talk will provide you with the answers and an understanding of why, when and how to use Differential Privacy in a meaningful way.