[D] Synthetic data generation techniques for data privacy

• SDGym

  1 242 7.8 Python

  Benchmarking synthetic data generation methods.

  

I would suggest starting with "differentially private synthetic data generation". These methods utilize differential privacy and mostly protect against membership inference attacks, are very popular in the ML/DL community. I would also suggest reading up on privacy preserving ML methods in general and adversarial attacks against them (membership inference, inversion, reconstruction, property inference), but if you're keen on reading some code, check out sd-gym: https://github.com/sdv-dev/SDGym. The authors have collected implementations for a lot of PPSDG methods. Also I strongly suggest reading McMahan's 2016 paper: https://arxiv.org/abs/1607.00133.

• InfluxDB

  www.influxdata.com featured

  Power Real-Time Data Analytics at Scale. Get real-time insights from all types of time series data with InfluxDB. Ingest, query, and analyze billions of data points in real-time with unbounded cardinality.

  

Suggest a related project