Postprocess parboost ensemble components

Share:

Description

Postprocess the submodels

Usage

1
2
postprocess(response, family, list_of_models, postprocessing,
  cores_cv = detectCores())

Arguments

response

The response vector.

family

Boosting family.

list_of_models

List of submodels.

postprocessing

Type of postprocessing.

Details

Either euqally weigh all submodels or perform (penalized) regression on them to determine the weights.

Value

Predictions of postprocessed ensemble and a list of mboost models

Author(s)

Ronert Obst