mlpack/mlpack

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

5703 stars · 1720 forks observed · 2026-08-28

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

mlpack is a fast, header-only C++ machine learning library built on Armadillo, ensmallen, and cereal, offering a wide range of algorithms for classification, regression, clustering, and nearest-neighbor search. It also ships command-line programs and bindings for Python, Julia, R, and Go.

Use cases

• train classifiers and regressors in C++
• run fast nearest-neighbor and range searches
• use machine learning from Python, R, Julia, or Go via bindings
• deploy lightweight ML models in production C++ apps
• cluster data with k-means or mean shift
• detect objects with YOLOv3
• cross-compile ML to embedded systems like Raspberry Pi

When to choose

• you need fast, lightweight C++ ML without heavy dependencies
• you want one library covering classification, regression, clustering, and tree-based spatial search
• you need ML bindings across multiple languages from a single core
• you target embedded or resource-constrained deployment

When to avoid

• you primarily want deep learning with GPU acceleration and autodiff (PyTorch, TensorFlow)
• you need a Python-first ecosystem with rich notebooks and tooling
• you prefer writing C++ over higher-level ML frameworks
• you need the very latest research models not yet implemented in mlpack

Facets

library · maturity stable

machine-learning deep-learning image-processing cli machine-learning data-science developer-tools cpp python cli cross-platform windows header-only bindings nearest-neighbor-search scientific-computing armadillo numfocus algorithms linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="mlpack/mlpack")