DOC improve documentation of NCR #1017
Conversation
|
@glemaitre ready for review |
doc/under_sampling.rst
Outdated
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^ | ||
|
|
||
| The :class:`NeighbourhoodCleaningRule` is another "cleaning" algorithm. It removes | ||
| samples from the majority class that are closest to the boundary with the minority |
There was a problem hiding this comment.
| samples from the majority class that are closest to the boundary with the minority | |
| samples from the majority class that are the closest to the boundary formed by the samples of the minority class |
There was a problem hiding this comment.
I don't totally understand this sentence. Let me try a modification in a new commit.
doc/under_sampling.rst
Outdated
|
|
||
| The :class:`NeighbourhoodCleaningRule` expands on the cleaning performed by | ||
| :class:`EditedNearestNeighbours` by eliminating additional majority class samples if | ||
| they are among the 3 closest neighbours of a sample from the minority class. |
There was a problem hiding this comment.
We have a parameter controlling the 3-NN.
| they are among the 3 closest neighbours of a sample from the minority class. | |
| they are among the :math:`N` closest neighbours (i.e. using the parameter `n_neighbours`) of a sample from the minority class. |
There was a problem hiding this comment.
Throughout the docs we are using K as the number of neighbours, not N. I guess the n in n_neighbours comes from n=number. I'd rather stick to K if that's alright with you, for consitency. I'll fix this in a separate commit.
There was a problem hiding this comment.
Actually, I removed this sentence altogether as per below suggestion.
| The procedure for the :class:`NeighbourhoodCleaningRule` is as follows: | ||
|
|
||
| 1. Remove observations from the majority class with edited nearest neighbors (ENN). | ||
| 2. Remove additional samples from the majority class if they are one of the k closest |
There was a problem hiding this comment.
Since we repeating the same sentence as above, I would remove the paragraph above and only go with the bullet point sequence.
doc/under_sampling.rst
Outdated
| To carry out step 2 there is one condition: a sample will only be removed if its class | ||
| has a minimum number of observations. The minimum number of observations is regulated | ||
| by the `threshold_cleaning` parameter. In the original article | ||
| :cite:`laurikkala2001improving`, samples would be removed if the class had at |
There was a problem hiding this comment.
I would not go in details regarding the original paper but instead just phrase that we check that the number of samples in the class to under-sample is above the threshold times the number of samples in the minority class.
imblearn/under_sampling/_prototype_selection/_neighbourhood_cleaning_rule.py
Outdated
Show resolved
Hide resolved
imblearn/under_sampling/_prototype_selection/_neighbourhood_cleaning_rule.py
Outdated
Show resolved
Hide resolved
glemaitre
changed the title
|
How can I check the linting error message? |
imblearn/under_sampling/_prototype_selection/_neighbourhood_cleaning_rule.py
Outdated
Show resolved
Hide resolved
Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
|
Thank you! |
Reword documentation and docstrings for the NCR.
Related to #854