Arturo Amor

Maybe briefly mention how data can affect these distributions (data imbalance, for instance)

Thanks for contributing @Asami-1! A first general comments regarding your PR: In general we try to avoid introducing external images to the documentation, instead we prefer adding references to existing...

This PR is good to be reviewed now.

For info this PR is failing on the CI with the following traceback: ```python-traceback File "/home/circleci/project/examples/miscellaneous/plot_outlier_detection_bench.py", line 218, in y_pred[model_name]["ames_housing"] = fit_predict( File "/home/circleci/project/examples/miscellaneous/plot_outlier_detection_bench.py", line 84, in fit_predict y_pred =...

The `[doc build]` directive does not seem to be building the documentation. Any suggestion on this @glemaitre?

> Is this the kind of output you expect? Yes, this is it. Thank you! Still would be nice to understand the building process, though :)

Below is the report of my tuning of the parameter `min_df`. My personal suggestion is to use `min_df=12`. WDYT? ---------------------- With `min_df=10` we have: - [Comparing RF and HGBT models](https://scikit-learn.org/stable/auto_examples/ensemble/plot_forest_hist_grad_boosting_comparison.html)...

> So not just 5th but you're also hiding 4th level like `Algorithm description` or `Mathematical details`? I am working on better guidelines for the use of dropdowns on #26636....

> Should we merge or we still need more reviews? @Charlie-XIAO For the dropdown PRs we typically require just one approve, so feel free to merge :)

Hi @chainesanbuenaventura, please let us know if you need help addressing the comments from the reviewers.