Adding documentation about related library: Concrete ML, for privacy preserving ML

[bcm-at-zama]

Describe the issue linked to the documentation

We would like to add Concrete ML in the documentation. Concrete ML is an open-source package, providing privacy-preserving ML, thanks to so-called fully homomorphic encryption. Concrete ML contains so-called built-in models which are very very close to scikit-learn, XGB and skorch. Eg, see https://docs.zama.ai/concrete-ml/built-in-models/linear

We've started to discuss with some of you, who advised me to go for an issue.

I would propose:

• [x] adding a category Privacy Preserving Machine Learning in ./doc/related_projects.rst, and adding Concrete ML inside --- done, this was https://github.com/scikit-learn/scikit-learn/pull/27376
• [x] adding a category Privacy Preserving Machine Learning in examples, and copy some of the examples of https://github.com/zama-ai/concrete-ml/tree/main/docs/advanced_examples, or make new ones (what you prefer!) [not to be done]
• [ ] maybe propose a blog post, if you guest us. Will be done soon, https://github.com/zama-ai/concrete-ml-internal/issues/4033

and we are actually open to lot of suggestions.

What do you think?

Suggest a potential alternative/fix

Adding some content as proposed in the description.

[adrinjalali]

We're not gonna depend on concrete_ml in our docs, so an example in the example gallery is not an option here. But we can review a blog post for our blog.

cc @scikit-learn/communication-team

[francoisgoupil]

We are working on a blog post with @bcm-at-zama

[bcm-at-zama]

Absolutely! It's https://github.com/zama-ai/concrete-ml-internal/issues/4033 in our internal repo. Stay tuned