SynthVAE
SynthVAE copied to clipboard
nhsx
Synthetic data generation by a Variational AutoEncoder with Differential Privacy assessed using Synthetic Data Vault metrics
Hello, I am trying to reproduce the distribution metrics established using SUPPORT, as stated on page 13 of the SynthVAE report. I have downloaded available code and checked that my...
**Describe the bug** A clear and concise description of what the bug is. **To Reproduce** Steps to reproduce the behavior: 1. Go to '...' 2. Click on '....' 3. Scroll...
**Describe the bug** When running the `GMM` preprocessing, we are seeing poor correlations coming out of the trained VAE. Investigate this further. **To Reproduce** Steps to reproduce the behavior: 1....
Currently privacy and fairness metrics are lacking - would be good to deepen the library of metrics available to the user such that they can get a complete breakdown of...
Currently SynthVAE cannot handle missing data in training sets - would be interesting in future to look at ways to amend this
allow user to adapt encoder/decoder structure e.g. 1D CNN architectures or deeper layers with different hidden layer dimensions
At the moment SDMETRICS gives many warnings originating from line 40 in their utils file https://github.com/sdv-dev/SDMetrics/blob/master/sdmetrics/utils.py. This function calculates frequencies of values in original data & synthetic data and throws...
We have suppressed warnings from SDV around a lot of datetimes not appearing in the original dataset - this seems to be due to the need to round the synthetic...