Related Projects

Here is a list of 3rd party open source projects who aim at following the same API conventions and coding guidelines (for instance the fit / predict / transform methods and the pervasive use of numpy arrays or scipy.sparse matrices as main data-structures.

Add your project here with the following infos:

• Project name and URL
• Short descriptive paragraph emphasizing implemented algorithms.
• License
• Link to the online documentation (if different from main site)
• Link to main source control repository
• scikit-learn compatibility status: are all the afore-mentioned public API implemented?

Optimum-Path Forest Classifier

LibOPF3 is a implementation of the Optimum-Path Forest general classifier, which aims for training speed and simplicity of params while keeping accuracy comparable to SVMs. It is licensed under the BSD license and presents both supervised and unsupervised version of the algorithm.

OPF is a general classifier which:

• supervised and unsupervised learning
• naturally multi-class
• fast fitting and predicting
• good accuracy
• few parameters
• allows some superposition between clusters

Link to Documentation, API and Repository

scikit-learn compatibility status: It implements a fit / predict interface and supports pickling (serialization)