[WIP] Add support for multilabel

[glemaitre]

Reference Issue

closes #340

What does this implement/fix? Explain your changes.

Any other comments?

[glemaitre]

@chkoar this is a partial solution to #340
However I get into trouble with the ensemble.
I would be happy any solution which can adapt to the output of any samplers (I think that I could iterate on the output of y in fact)

[codecov]

Codecov Report

Merging #341 into master will decrease coverage by 0.19%.
The diff coverage is 97.43%.

@@            Coverage Diff            @@
##           master     #341     +/-   ##
=========================================
- Coverage   98.78%   98.59%   -0.2%     
=========================================
  Files          68       68             
  Lines        3961     3988     +27     
=========================================
+ Hits         3913     3932     +19     
- Misses         48       56      +8

Impacted Files	Coverage Δ
imblearn/combine/smote_tomek.py	100% <100%> (ø)	⬆️
imblearn/ensemble/balance_cascade.py	100% <100%> (ø)	⬆️
imblearn/utils/estimator_checks.py	96.25% <100%> (+0.42%)	⬆️
imblearn/combine/smote_enn.py	100% <100%> (ø)	⬆️
imblearn/utils/validation.py	99.18% <92.3%> (-0.82%)	⬇️
imblearn/under_sampling/base.py	98.11% <0%> (-1.89%)	⬇️
imblearn/tests/test_common.py	94.73% <0%> (-0.72%)	⬇️
...arn/under_sampling/prototype_selection/nearmiss.py	98.55% <0%> (-0.08%)	⬇️
imblearn/over_sampling/smote.py	92.9% <0%> (-0.05%)	⬇️
imblearn/over_sampling/adasyn.py	98.48% <0%> (-0.03%)	⬇️
... and 31 more

---

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 23ad602...b445cb8. Read the comment docs.