MAINT deprecate n_jobs in over-sampling algorithms

[glemaitre]

Most of the time n_jobs is not used anymore. Currently it was really useful to use it in scikit-learn and it can be set directly by passing an estimator with the parameter set which is a better choice.

[codecov]

Codecov Report

Merging #887 (11f5031) into master (51c56a1) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master     #887   +/-   ##
=======================================
  Coverage   97.31%   97.32%           
=======================================
  Files          97       97           
  Lines        6367     6389   +22     
  Branches      720      727    +7     
=======================================
+ Hits         6196     6218   +22     
  Misses        102      102           
  Partials       69       69           

Impacted Files	Coverage Δ
imblearn/over_sampling/_adasyn.py	88.05% <100.00%> (+0.55%)	⬆️
imblearn/over_sampling/_smote/base.py	99.01% <100.00%> (+0.03%)	⬆️
imblearn/over_sampling/_smote/filter.py	88.33% <100.00%> (+0.50%)	⬆️
imblearn/over_sampling/tests/test_common.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 51c56a1...11f5031. Read the comment docs.