---
id: scikit-multilearn
version: "0.2.0"
license: BSD
license_treatment: permissive
maintenance: abandoned
---

# scikit-multilearn — Scikit-multilearn is a BSD-licensed library for multi-label classification that is built on top of the well-known scikit-learn ecosystem.

License: permissive · Maintenance: abandoned · Downloads: 77.7K/mo

## What it is and what it does

Scikit-multilearn is a Python library for multi-label classification—problems where each sample can belong to multiple classes simultaneously. It implements problem-transformation methods and algorithm-adaptation approaches, built on top of numpy and scipy with an API modeled after scikit-learn. The library also provides a wrapper for MEKA, MULAN, and WEKA, and includes community-detection-based clusterers for label correlation.

The project has been abandoned since 2018-12-10 with no active maintenance, meaning it will not receive updates for compatibility with newer Python or dependency versions, nor any security patches. Installation is low-friction with pure Python wheels, but long-term viability for new projects is questionable.

Use it for:

- Text categorization where documents can belong to multiple topics simultaneously.
- Image tagging where a single image can have multiple relevant labels or tags.
- Biological sequence annotation where genes or proteins may have multiple functional roles.
- Recommendation systems where items can belong to multiple categories or user interests.
- Medical diagnosis where a patient may have multiple concurrent conditions.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Scikit-multilearn performs multi-label classification tasks using problem-transformation and algorithm-adaptation methods, built on numpy and scikit-learn with a compatible API.

No, unless you are maintaining legacy code. The package is abandoned (last release 2018-12-10) and will not receive updates for modern Python or dependency versions. For new multi-label classification projects, consider actively maintained alternatives. If you must use it, pin all dependencies to versions known to work with scikit-multilearn 0.2.0 and test thoroughly.

## Install

pip install scikit-multilearn
uv add scikit-multilearn
poetry add scikit-multilearn

## Installing scikit-multilearn

Before you install: Installation is low-friction with pure Python wheels available. However, the package is abandoned—last release was 2018-12-10. No active maintenance or security updates should be expected.

License in practice: BSD license is permissive and poses no restrictions on commercial or private use, modification, or redistribution.

Quickstart:

pip install scikit-multilearn

from skmultilearn.problem_transform import BinaryRelevance

classifier = BinaryRelevance(require_dense=[False,True])
classifier.fit(X_train, y_train)
y_pred = classifier.predict(X_test)

Requires numpy, scipy, scikit-learn, and other dependencies; some optional clusterers require GPL-licensed packages.

Verify before relying:

- Whether the package works with current versions of scikit-learn and numpy given its 2018 release date.
- Compatibility with modern Python versions not specified in the fact sheet.
- Whether the MEKA wrapper and other integrations remain functional without active maintenance.
- Current state of optional GPL-licensed dependencies (python-igraph, python-graphtool) and their installation complexity.

## Package facts

- License: BSD (permissive)
- Python support: unspecified
- Install friction: low
- Maintenance: abandoned
- Downloads: 77.7K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

multi-label classification, multilabel learning python, multiple labels per sample, scikit-learn multi-label, binary relevance classification, label powerset, multi-label ensemble methods, multi-label-learning, abandoned-package

[View on SkillFed](https://skillfed.io/packages/scikit-multilearn) · [View on PyPI](https://pypi.org/project/scikit-multilearn/)