Create setting for multi-resolution CovNN model (stucture based on https://arxiv.org/pdf/1608.00647.pdf CNN1)

setCovNN(batchSize = 1000, outcomeWeight = 1, lr = 1e-05, decay = 1e-06,
  dropout = 0, epochs = 10, filters = 3, kernelSize = 10,
  loss = "binary_crossentropy", seed = NULL)

Arguments

batchSize

The number of samples to used in each batch during model training

outcomeWeight

The weight assined to the outcome (make greater than 1 to reduce unballanced label issue)

lr

The learning rate

decay

The decay of the learning rate

dropout

[currently not used] the dropout rate for regularisation

epochs

The number of times data is used to train the model (e.g., epoches=1 means data only used once to train)

filters

The number of columns output by each convolution

kernelSize

The number of time dimensions used for each convolution

loss

The loss function implemented

seed

The random seed

Examples

# NOT RUN {
model.CovNN <- setCovNN()
# }