make_ensemble: Build a weighted ensemble list

Description Usage Arguments Value

Description

Gather all parameters of a weighted ensemble of RWeka models into formated list

Usage

1
2
3
make_ensemble(train_preds = NULL, model_list = NULL, weightType = NULL,
  comb_rule = NULL, bin_type = "standard", bin_features = NULL,
  nbins = NULL, knn_size = 10, rotate = FALSE)

Arguments

model_list

A list of pre trained models (using the same data as train)

weightType

How the ensemble should be weighted; "unweighted", "weighted", "bin weighted"

comb_rule

How the ensemble predictions should be comined; "average posterior","majority vote"

bin_type

How bins should be created when weightType is "bin weighted"; "average posterior","majority vote"

bin_features

The name of two numeric bins when weightType is "bin weighted"

nbins

The number of bins to create when weightType is "bin weighted"

train_data

Data on which to train the ensemble

Value

A trained ensemble that can be used to predict new data points


kmaurer/binsemble documentation built on May 7, 2019, 9:50 p.m.