API Reference¶

Extreme Learning Machine is a native regression method, that also supports classification with extra features designed to support classes in targets.

Large ELM refers to a parallel implementation with Dask that takes data in a bunch of files on disk, and supports extremely large numbers of neurons - at a cost of slightly longer processing times.

Hidden layer, solvers and utilities are references to internal implementations for those interested in extending or re-using parts of Scikit-ELM.

Regressor¶

`ELMRegressor`([alpha, batch_size, ...])	Extreme Learning Machine for regression problems.
`LargeELMRegressor`(SLFNs, solver)	ELM Regressor for larger-than-memory problems.

Classifier¶

ELMClassifier([classes, alpha, batch_size, ...])

ELM classifier, modified for multi-label classification support.

Hidden Layer¶

HiddenLayer([n_neurons, density, ufunc, ...])

Scikit-Learn compatible interface for SLFN.

Solvers¶

BatchCholeskySolver([alpha])

Utilities¶

utils