Assessment of Neural Network Set Up — assessAutoencoderOptions • SPATA2

Assesses different neural network set ups regarding the activation function and the number of bottleneck neurons.

assessAutoencoderOptions(
  expr_mtr,
  activations,
  bottlenecks,
  layers = c(128, 64, 32),
  dropout = 0.1,
  epochs = 20,
  verbose = TRUE
)

runAutoencoderAssessment(
  object,
  activations,
  bottlenecks,
  layers = c(128, 64, 32),
  dropout = 0.1,
  epochs = 20,
  verbose = TRUE
)

Arguments

expr_mtr: The expression matrix that is to be used as input for the neural network.
activations: Character vector. Denotes the activation functions to be assessed.
bottlenecks: Numeric vector. Denotes the different numbers of bottleneck neurons to be assessed.
layers: Numeric vector of length 3. Denotes the number of neurons in the three hidden layers. (default = c(128, 64, 32))
dropout: Numeric value. Denotes the dropout. (defaults to 0.1)
epochs: Numeric value. Denotes the epochs of the neural network. (defaults to 20)
object: A valid spata-object.

Value

runAutoencoderAssessment(): The spata object containing the list that holds the total variance measured by irlba::prcomp_irlba() after each combination of activations/bottlenecks as well as the additional set up.
assessAutoencoderOptions(): The list that holds the total variance measured by irlba::prcomp_irlba() after each combination of activations/bottlenecks as well as the additional set up.