imbalanced-learn

imbalanced-learn is a python package offering a number of re-sampling techniques commonly used in datasets showing strong between-class imbalance. It is compatible with scikit-learn and has been submitted to be part of scikit-learn-contrib projects.

Documentation

Installation documentation, API documentation, and examples can be found on the documentation.

Installation

Dependencies

imbalanced-learn is tested to work under Python 2.7 and Python 3.5.

scipy(>=0.17.0)
numpy(>=1.10.4)
scikit-learn(>=0.17.1)

Installation

imbalanced-learn is currently available on the PyPi’s reporitories and you can install it via pip:

pip install -U imbalanced-learn

The package is release also in Anaconda Cloud platform:

conda install -c glemaitre imbalanced-learn

If you prefer, you can clone it and run the setup.py file. Use the following commands to get a copy from Github and install all dependencies:

git clone https://github.com/fmfn/UnbalancedDataset.git
cd UnbalancedDataset
python setup.py install

Testing

After installation, you can use nose to run the test suite:

make coverage

About

Most classification algorithms will only perform optimally when the number of samples of each class is roughly the same. Highly skewed datasets, where the minority is heavily outnumbered by one or more classes, have proven to be a challenge while at the same time becoming more and more common.

One way of addresing this issue is by re-sampling the dataset as to offset this imbalance with the hope of arriving at a more robust and fair decision boundary than you would otherwise.

Re-sampling techniques are divided in two categories:

Under-sampling the majority class(es).
Over-sampling the minority class.
Combining over- and under-sampling.
Create ensemble balanced sets.

Below is a list of the methods currently implemented in this module.

Under-sampling
1. Random majority under-sampling with replacement
2. Extraction of majority-minority Tomek links [1]
3. Under-sampling with Cluster Centroids
4. NearMiss-(1 & 2 & 3) [2]
5. Condensend Nearest Neighbour [3]
6. One-Sided Selection [4]
7. Neighboorhood Cleaning Rule [5]
8. Edited Nearest Neighbours [6]
9. Instance Hardness Threshold [7]
10. Repeated Edited Nearest Neighbours [14]
Over-sampling
1. Random minority over-sampling with replacement
2. SMOTE - Synthetic Minority Over-sampling Technique [8]
3. bSMOTE(1 & 2) - Borderline SMOTE of types 1 and 2 [9]
4. SVM SMOTE - Support Vectors SMOTE [10]
5. ADASYN - Adaptive synthetic sampling approach for imbalanced learning [15]
Over-sampling followed by under-sampling
1. SMOTE + Tomek links [12]
2. SMOTE + ENN [11]
Ensemble sampling
1. EasyEnsemble [13]
2. BalanceCascade [13]

The different algorithms are presented in the following notebook.

This is a work in progress. Any comments, suggestions or corrections are welcome.

References:

Metadata

Release files for imbalanced-learn 0.1.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for imbalanced-learn 0.1.1
File	Size	Uploaded
imbalanced-learn-0.1.1.tar.gz	5.0 MB	Jul 9, 2016	Details

Built distribution (wheel)

Table of built distributions (wheels) for imbalanced-learn 0.1.1
File	Interpreter	ABI	Platform	Reset
imbalanced_learn-0.1.1-py2-none-any.whl 89.6 kB Jul 9, 2016	Python 2	none	any	Details

Total release size: 5.1 MB

Release files / imbalanced-learn-0.1.1.tar.gz

Download URL	imbalanced-learn-0.1.1.tar.gz
Size	5.0 MB
Tags	Source
SHA-256 checksum How to use checksums	`015ab121937d8bcb7654f07d6e20509c839f9cc6df48740f5d5fa94aec9b8bfc`
BLAKE2b-256 checksum How to use checksums	`c2f0542179bdf2aae2242c56cd91d5e98803152cd3a116fd989c63e234db830f`
Upload date	Jul 9, 2016
Uploaded using Trusted Publishing? What is trusted publishing?	No

Release files / imbalanced_learn-0.1.1-py2-none-any.whl

Download URL	imbalanced_learn-0.1.1-py2-none-any.whl
Size	89.6 kB
Tags	Python 2
SHA-256 checksum How to use checksums	`04eb83cf42c4f3782cd7fc9eef5b6bfcbf651490f9c8d3e13ddfefa9b5415c08`
BLAKE2b-256 checksum How to use checksums	`4c6dfd09946d21b3b31c3093352384fc70121d9a46b91f0ff572ef606c18121d`
Upload date	Jul 9, 2016
Uploaded using Trusted Publishing? What is trusted publishing?	No

Release history Release notifications | RSS feed

0.14.2

Jun 7, 2026 2 release files

0.14.1

Dec 21, 2025 2 release files

0.14.0

Aug 14, 2025 2 release files

0.13.0

Dec 20, 2024 2 release files

0.12.4

Oct 4, 2024 2 release files

0.12.3

May 28, 2024 2 release files

0.12.2

Mar 31, 2024 2 release files

0.12.1

Mar 31, 2024 2 release files

0.12.0

Jan 24, 2024 2 release files

0.11.0

Jul 8, 2023 2 release files

0.10.1

Dec 28, 2022 2 release files

0.10.0

Dec 9, 2022 2 release files

0.9.1

May 16, 2022 2 release files

0.9.0

Jan 11, 2022 2 release files

0.8.1

Sep 29, 2021 2 release files

0.8.0

Feb 18, 2021 2 release files

0.7.0

Jun 9, 2020 2 release files

0.6.2

Feb 16, 2020 2 release files

0.6.1

Dec 7, 2019 2 release files

0.6.0

Dec 5, 2019 2 release files

0.5.0

Jun 28, 2019 2 release files

0.4.3

Nov 6, 2018 2 release files

0.4.2

Oct 21, 2018 2 release files

0.4.1

Oct 12, 2018 2 release files

0.4.0

Oct 12, 2018 2 release files

0.3.3

Feb 22, 2018 2 release files

0.3.2

Dec 7, 2017 2 release files

0.3.1

Oct 9, 2017 2 release files

0.3.0

Aug 24, 2017 3 release files

0.2.1

Jan 1, 2017 2 release files

0.2.0

Dec 31, 2016 2 release files

0.1.9

Dec 26, 2016 2 release files

0.1.8

Sep 7, 2016 2 release files

0.1.7

Aug 31, 2016 2 release files

0.1.6

Aug 9, 2016 2 release files

0.1.5

Jul 31, 2016 2 release files

0.1.3

Jul 19, 2016 2 release files

0.1.2

Jul 19, 2016 2 release files

This release

0.1.1 This release

Jul 9, 2016 2 release files

0.1

Jul 6, 2016

imbalanced-learn 0.1.1