DOC Remove examples for the old tutorials section #30460
Conversation
There was a problem hiding this comment.
Maybe @lesteve or @glemaitre may be interested in looking at this?
| This exercise is used in the :ref:`cv_estimators_tut` part of the | ||
| :ref:`model_selection_tut` section of the :ref:`stat_learn_tut_index`. |
There was a problem hiding this comment.
This references the old doc/tutorial/statistical_inference/model_selection.rst (you can see it in #29104).
I think the user guide section https://scikit-learn.org/stable/modules/grid_search.html#model-specific-cross-validation pretty much gives similar information as the old tutorial. I would add a link to this example there, but I am a little confused about what this example is trying to say? That GridSearchCV cannot be trusted?!
|
|
||
| This exercise is used in the :ref:`clf_tut` part of the |
There was a problem hiding this comment.
This references the old doc/tutorial/statistical_inference/supervised_learning.rst:
For classification, as in the labeling
`iris <https://en.wikipedia.org/wiki/Iris_flower_data_set>`_ task, linear
regression is not the right approach as it will give too much weight to
data far from the decision frontier. A linear approach is to fit a sigmoid
function or **logistic** function:For classification, as in the labeling
`iris <https://en.wikipedia.org/wiki/Iris_flower_data_set>`_ task, linear
regression is not the right approach as it will give too much weight to
data far from the decision frontier. A linear approach is to fit a sigmoid
function or **logistic** function:I am actually not clear what " linear regression is not the right approach as it will give too much weight to data far from the decision frontier" means? There is not much text here so I am not exactly sure what it is conveying.
| @@ -4,10 +4,6 @@ | |||
| ================================ | |||
|
|
|||
| A tutorial exercise for using different SVM kernels. | |||
|
|
|||
| This exercise is used in the :ref:`using_kernels_tut` part of the | |||
There was a problem hiding this comment.
This links to the old doc/tutorial/statistical_inference/supervised_learning.rst, which starts off with:
Classes are not always linearly separable in feature space. The solution is to
build a decision function that is not linear but may be polynomial instead.
This is done using the *kernel trick* that can be seen as
creating a decision energy by positioning *kernels* on observations:which is nice and explicit, and I don't think the current user guide kernel section does this: https://scikit-learn.org/stable/modules/kernel_approximation.html#kernel-approximation . Would be a separate issue, but we could consider adding bits from this old tutorial page.
Not much text here, I think it is just to show how to use a kernel? I am not sure if there is anywhere in the current user guide that should link to this.
|
Honestly those were exercises for the old tutorials. Since old tutorials were removed I would say we can remove the |
|
Yeah I would agree with you. I've amended this PR to now delete the tutorial examples. Let's see what others think now. |
Reference Issues/PRs
Inspired after #30407 (comment)
What does this implement/fix? Explain your changes.
Currently docs has the following warnings:
Warnings
The tutorials section was removed #29104, the examples gallery
examples/exercisesstill references the old tutorials section. These examples are not very verbose and I don't think they add much value so this PR deletes themAny other comments?