feat: Add some description helpers to the EstimatorReport

[sylvaincom]

Is your feature request related to a problem? Please describe.

This feature request is an improvement of the EstimatorReport of skore v0.6.1.

As a data scientist, when using the EstimatorReport on a classifier, I would like to get some quick information of what is the accuracy, precision, recall, and F1-score, in case I forgot exactly what is the difference between precision and recall, and how to interpret them.

I did a user interview with a relatively junior data scientist who had not coded for around a year. He knew that precision and recall were scores, but he forgot the exact formula and how to interpret them. We should add some quick cheat sheets on that, without the need to read the documentation.

Describe the solution you'd like

I would like for skore to have something similar to skrub's TableReport, in the "Associations" tab, with a short description on how to interpret the association with an external link for more information:

Note that we could also provide a link to Vinny's videos:

• Precision, recall and F1-score
• When precision equals recall

Describe alternatives you've considered, if relevant

No response

Additional context

No response

[glemaitre]

I think that it will be difficult in the "rich" repr. However, with an improve HTML representation, we could add a (?) link next to the metric that could redirect to the scikit-learn documentation. A bit like for the sckit-learn estimator diagram:

[sylvaincom]

Sure, looks great