LearnBase

WORK IN PROGRESS

This package is an attempt to provide common interfaces and function definitions for Machine Learning packages in Julia

Everything here is subject to change. The initial code here is just factored out code that I have currently in use.

Example

Common class encodings for machine learning algorithms that need numeric target vectors

ZeroOneClassEncoding,
SignedClassEncoding,
MultivalueClassEncoding,
OneOfKClassEncoding

Abstract types for convenience. Extending from them instead of the base types will automatically take care of non-numeric target vectors using a decorator Their behavious is very similar to DataFrameModels do.

abstract EncodedStatisticalModel{E<:ClassEncoding} <: StatisticalModel
abstract EncodedRegressionModel{E<:ClassEncoding} <: RegressionModel