How do you generate fake data that closely resembles your original data? This post discusses one approach to do it.

1

This post discusses differential privacy as a way of thinking about machine unlearning

2