Tune a deep learning model

Usage

tl_tune_deep(
  data,
  formula,
  is_classification = NULL,
  hidden_layers_options = list(c(32), c(64, 32), c(128, 64, 32)),
  learning_rates = c(0.01, 0.001, 1e-04),
  batch_sizes = c(16, 32, 64),
  epochs = 30,
  validation_split = 0.2,
  ...
)

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. NULL (default) reads it from the response, as tl_model does: a factor or character response is classification. FALSE with such a response is an error.
hidden_layers_options: List of vectors defining hidden layer configurations to try
learning_rates: Learning rates to try (default: c(0.01, 0.001, 0.0001))
batch_sizes: Batch sizes to try (default: c(16, 32, 64))
epochs: Number of training epochs (default: 30)
validation_split: Proportion of the rows held out to score each configuration on, drawn at random (default: 0.2). Every configuration is scored on the same rows.
...: Additional arguments passed to keras's fit() for every configuration; verbose (default 0) replaces the value used otherwise. Arguments with one value per row – weights, subset, offset, foldid, strata – are refused, since each configuration is fitted on part of the rows.

Value

A list with elements model (the best configuration refitted as a tidylearn_model, so predict() and the deep plots take it; the keras model is at $model$fit$model), best_hidden_layers (optimal layer configuration), best_learning_rate, best_batch_size, and tuning_results (a data frame of all hyperparameter combinations and their validation losses).

Examples

if (FALSE) { # \dontrun{
if (requireNamespace("keras", quietly = TRUE)) {
  result <- tl_tune_deep(iris, Species ~ .,
    hidden_layers_options = list(c(10), c(10, 5)),
    learning_rates = c(0.01, 0.001), batch_sizes = c(32),
    epochs = 5)
  predict(result$model, iris[1:5, ])
}
} # }