Example of encrypted learning with partial_fit models
- Dominant language
- Python
- Stars
- 951
- Forks
- 262
- PR merge metrics
- No merged PRs in 30d
Description
Hi developers, I am unfortunately not in the position to write it myself due to other obligations.
Despite that, I wanted to share the tutorial below I came across on HN. Perhaps someone is interested in recreating this tutorial with the `partial_fit ` methods as used in http://ml.dask.org/incremental.html
I think it's a great example that showcases the power of incremental learning models in a time where Privacy is getting more and more relevant.
https://github.com/OpenMined/PySyft/tree/master/examples/tutorials
Kind regards.
Contributor guide
Research direction
Review the linked PySyft examples/tutorials and the Dask-ML incremental-learning page first; the issue does not identify a target file or test. Define the encrypted partial_fit example's scope and location, then verify that the completed tutorial reproducibly demonstrates encrypted incremental learning.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- cryptography, machine-learning
- Issue type
- Documentation
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100