scikit-learn-contrib/imbalanced-learn

Health v2 · maintenance only

Adoption not part of the score

7121 stars · 1360 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

imbalanced-learn is a Python library offering re-sampling techniques (over-sampling, under-sampling, and combinations) for datasets with strong class imbalance. It is fully compatible with scikit-learn and part of the scikit-learn-contrib ecosystem.

Use cases

• handle imbalanced classification datasets
• apply SMOTE oversampling in Python
• undersample majority class in training data
• resample data before training a scikit-learn model
• fix class imbalance in fraud or anomaly detection datasets
• combine over- and under-sampling pipelines

When to choose

• your classification dataset has heavily skewed class distributions
• you want scikit-learn-compatible transformers and pipelines for resampling
• you need well-established techniques like SMOTE, ADASYN, or Tomek links

When to avoid

• your dataset is balanced or only mildly imbalanced
• you need deep-learning-specific imbalance handling outside the scikit-learn ecosystem
• you need anomaly detection rather than supervised resampling

Facets

library · maturity stable

machine-learning data-science testing machine-learning data-science python cross-platform imbalanced-data resampling smote scikit-learn classification

Member repositories

For agents

markdown · JSON · MCP: product_card(name="scikit-learn-contrib/imbalanced-learn")