Build database using data synthesis

[yoid2000]

One way of produce anonymized data is to build a synthesized data set from the raw data. This process, so far as I understand it, normally involves building a statistical model of the data, and then generating data from the statistical model. Such an approach may be particularly good for subsequent use in ML.

I'm not very familiar with this, so don't have any suggestions for starting points. Will need some research on existing tools to start.