Skip to contents

Compute Cross-Validated AUC Values for Machine Learning Models

Source: R/machine_learning.R
compute_cv_AUC.Rd

This function computes cross-validated AUROC and AUPRC scores for a list of trained machine learning models. It can also save performance barplots and optionally select base models for stacking.

Usage

compute_cv_AUC(
  models,
  file_name = NULL,
  base_models = FALSE,
  AUC_type = "AUROC",
  return = TRUE
)

Arguments

models

A named list of trained machine learning models. Each model should contain a resample data frame with AUROC and AUPRC values from cross-validation.

file_name

(Optional) Character string. Used as the prefix for the plot filenames if save_plot = TRUE.

base_models

Logical. If TRUE, selects a subset of models as base learners for stacking using the choose_base_models() function.

AUC_type

Character. Either "AUROC" or "AUPRC"; determines which metric is used to select the top-performing model.

return

Logical. Whether to return the results and generated plots.

Value

A list containing:

AUROC

A data frame with median and standard deviation of AUROC values for each model.

AUPRC

A data frame with median and standard deviation of AUPRC values for each model.

Top_model

The name of the model with the highest median value for the selected metric (AUC_type).

Base_models

(Optional) A character vector of selected base models for stacking, returned if base_models = TRUE.

Examples

if (FALSE) { # \dontrun{
res <- compute_cv_AUC(
  models = ml_models,
  file_name = "Model_Performance",
  base_models = TRUE,
  AUC_type = "AUROC",
  save_plot = TRUE
)
} # }