Tune a neural network model

Usage

tl_tune_nn(
  data,
  formula,
  is_classification = FALSE,
  sizes = c(1, 2, 5, 10),
  decays = c(0, 0.001, 0.01, 0.1),
  folds = 5,
  ...
)

Arguments

data: A data frame containing the training data
formula: A formula specifying the model
is_classification: Logical indicating if this is a classification problem
sizes: Vector of hidden layer sizes to try
decays: Vector of weight decay parameters to try
folds: Number of cross-validation folds (default: 5)
...: Additional arguments to pass to nnet()

Value

A list with elements model (the best fitted nnet model), best_size (optimal hidden-layer size), best_decay (optimal weight decay), and tuning_results (a data frame of all parameter combinations and their cross-validated errors).

Examples

# \donttest{
tuned <- tl_tune_nn(iris, Species ~ .,
  is_classification = TRUE,
  sizes = c(2, 5), decays = c(0, 0.01), folds = 3)

tuned$best_size
#> [1] 5
tuned$best_decay
#> [1] 0
tuned$tuning_results
#>   size decay      error
#> 1    2  0.00 0.15333333
#> 2    5  0.00 0.02000000
#> 3    2  0.01 0.02000000
#> 4    5  0.01 0.03333333

# The grid this searched, drawn as a heatmap
tl_plot_nn_tuning(tuned)

# }