LogisticRegressionLearner: Remove deprecated argument 'multi_class'

[janezd]

Issue

As per https://scikit-learn.org/1.5/modules/generated/sklearn.linear_model.LogisticRegression.html,

Deprecated since version 1.5: multi_class was deprecated in version 1.5 and will be
removed in 1.7. From then on, the recommended ‘multinomial’ will always be used
for n_classes >= 3. Solvers that do not support ‘multinomial’ will raise an error. Use
sklearn.multiclass.OneVsRestClassifier(LogisticRegression()) if you still want to use OvR.

Description of changes

Remove the argument.

Includes

• Code changes
• Tests
• Documentation

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 88.47%. Comparing base (c2c1648) to head (57f85cc).
Report is 19 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #6956      +/-   ##
==========================================
+ Coverage   88.45%   88.47%   +0.02%     
==========================================
  Files         331      332       +1     
  Lines       73099    73277     +178     
==========================================
+ Hits        64659    64834     +175     
- Misses       8440     8443       +3     

[markotoplak]

In _initialize_wrapped we'd also have to switch the l1 solver to "saga", because "liblinear" does not support "multiclass".