OpenMined / PyDP

The Python Differential Privacy Library. Built on top of: https://github.com/google/differential-privacy
Apache License 2.0
503 stars 139 forks source link

Research on IBM-DP #302

Open chinmayshah99 opened 4 years ago

chinmayshah99 commented 4 years ago

What?

As part of our new road-map, we want to integrate other noise mechanisms, and IBMDP has a lot of functionality from noise mechanisms to ML algorithms.

The research would include how the noise mechanisms works+ ML algorithms, what all does that library support and how can we integrate it with PyDP.

Note: This is open-ended research and the scope of questions might not be limited to mentioned above.

How long?

2 weeks

mpcsb commented 3 years ago

I can take this one. :)

chinmayshah99 commented 3 years ago

Sure

dnabanita7 commented 3 years ago

Is it still available? seems interesting!

mpcsb commented 3 years ago

It is. I've let this slide. Thanks

chinmayshah99 commented 3 years ago

Hi @dnabanita7 , the issue is still available. Though I would strongly encourage you to discuss this in #lib_pydp channel first.

dnabanita7 commented 3 years ago

Hi @chinmayshah99 Should I discuss in #support channel?

dnabanita7 commented 3 years ago

Updates: Here is the link for the research done. Implementation is left.