ML.Local.lm: ML.Local.lm

Description Usage Format Methods

Description

Initializes a linear model (glm) for locally training an estimator. This uses the normal GLM implementation as provided by the R stats package.

Usage

1

Format

An object of class R6ClassGenerator of length 24.

Methods

initialize(learning.rate = 0.1, family = "binomial", initialization.random = FALSE)

Initializes a new generalized linear model to use in the online super learner.

@param learning.rate double (default = 0.1) the learning rate to use when training the model.

@param family string (default = "binomial") the family to use for the estimator. Generally this should be binomial (since we are running various logistic regressions). However, if you have a good reason to use a different specification, change it here.

@param initialization.random (default = FALSE) each of the parameters needs to have an intial state. If this argument is true, all parameters will be randomly initialized.

@param verbose (default = FALSE) the verbosity of the fitting procedure

get_validity

Active Method. Determines whether the current specification is valid or not. Note that this function does not return anything, instead it throws whenever something is incorrect.


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