rf.LOO: Random Forests with Leave-One-Out Procedures

In raytut/randomForestLOO: Random Forests with Leave-One-Out

Random Forests with Leave-One-Out Procedures

Description

Function to run a Leave-One-Out random forest model. Two methods can be specified: LOSO and LOBO.

LOSO (Leave-One-Subject-Out) will remove one entire subject from the training set, and then test the model on the removed subject, repeating on all subjects in the dataframe. This offers population level classification.

LOBO (Leave-One-Beep-Out) offers within subject level predictions for EMA(Ecological Momentary Assessment) type data by training models on single subjects minus one Beep. Classification are thus derived for each subject this way. This analysis can be implemented on other types of trial based data.

Usage

rf.LOO(data, sub_id, xvars, yvar, ntree, method, progress)

Arguments

data	Dataframe to be used by model
sub_id	String indicating what variable to use as subject ID
xvars	Vector of strings to be used for model training
yvar	A response string. Must be a character/factor. Otherwise, use rf.LOOreg for regression version (in development).
ntree	Integer indicating number of trees to simulate. Defaults to 1000.
method	String, can be "LOSO" or "LOBO". Defaults to "LOBO".
progress	Logical indicating if progress to be printed. Default True.

Value

Results of the LOO analysis as a list:

model	Information on the model that was run
subject	Results for each subject, including the predictions made as a list, a confusion matrix, and the mean error
errors	Mean error rate returned per subject
mean_errors	Population level error rate across subjects

Note

Currenlty implemented only for factors, and tested for two and three levels.

Author(s)

Rayyan Tutunji | rayyan.tutunji[at]donders.ru.nl

References

Using wearable biosensors and ecological momentary assessments for the detection of prolonged stress in real life Rayyan Tutunji, Nikos Kogias, Bob Kapteijns, Martin Krentz, Florian Krause, Eliana Vassena, Erno Hermans bioRxiv 2021.06.29.450360; doi: https://doi.org/10.1101/2021.06.29.450360

raytut/randomForestLOO documentation built on May 30, 2022, 8:47 p.m.