fit.OnlineSuperLearner: fit.OnlineSuperLearner

In frbl/OnlineSuperLearner: Online SuperLearner package

Description

Fits an online superlearner using a similar notation as a GLM.

Usage

fit.OnlineSuperLearner(formulae, data, algorithms = NULL,
  bounds = FALSE, ...)

Arguments

formulae	list a list of all relevantVariable objects that need to be fitted
data	data.frame or list of data.frames the data set to use for fitting the OSL
algorithms	list of algorithms to use in the online superlearner
bounds	either a list of bounds, or a boolean (default = FALSE), in which TRUE forces the bounds to be generated automatically, FALSE causes the bounds not to be generated at all (no normalization) we provide the option to normalize the data in the OSL procedure. This entails that the package will automatically select a set of bounds (min and max) based on the data set provided. After that it will only use the normalized features (all scaled between 0-1). The bounds should be specified as a list in which each element represents one of the RelevantVariable objects. Each of these entries should contain another list with two elements: min_bound and max_bound, which represent the lower and upper bound of that variable in specific.
...	other parameters directly passed to the OSL and fit function. There are several named variables to provide here: - initial_data_size - max_iterations - mini_batch_size See for a full list the documentation of the OnlineSuperLearner fit and initialize functions.

Value

a fitted version of an OnlineSuperLearner class

frbl/OnlineSuperLearner documentation built on Feb. 9, 2020, 9:28 p.m.