Skip to content

README.md

Latest commit

 

History

History
8 lines (5 loc) · 448 Bytes

README.md

File metadata and controls

8 lines (5 loc) · 448 Bytes

Pytorch Data Provider for Generating Artificially-Missing Data

The repository is currently work in progress. Its purpose is to generate synthetic datasets with missing values.

Future work

  • Implement Ampute algorithm from "Generating missing values for simulation purposes: a multivariate amputation procedure" by Schouten et al. (2018)
  • Add additional MNAR tests, e.g. KL-div of a model fitted on observed data and complete data.