runCLR: Optimizes the cyclical learning rate (CLR) based...

In kristianHoden/smartPARE: An R package for efficient identification of true mRNA cleavage sites

Description

The function optimizes the cyclical learning rate (CLR) based convolutional neural network (CNN) model based on the hyperparameters assigned

Usage

runCLR(
  optimizer2 = keras::optimizer_sgd(lr = lr_max1, decay = 0),
  denseLoop2,
  trainLabels2 = trainLabels,
  epochs2 = 5,
  batch_size2,
  dropout2 = 0.2,
  patience2 = 30,
  validation_split2 = 0.2,
  convolutionalLoop2,
  NO_pooling2 = 1,
  testPics2 = testPics,
  Learning_rate_l2 = Learning_rate_l,
  Learning_rate_h2 = Learning_rate_h,
  pathOut = paste0(homePath1, "/bayesmodels/")
)

Arguments

optimizer2	Optimizer
denseLoop2	Number of times to iterate the dense layers loops
trainLabels2	trainLabels object generated by
epochs2	Number of epochs in keras
batch_size2	Batch size in keras
dropout2	Keras dropout
patience2	EarlyStopping patience
validation_split2	Fraction of the dataset that will be used for validation
convolutionalLoop2	Number of times to iterate the convolutional layers loop
NO_pooling2	Number of times to perform pooling in the convolutional layers loop (max = 2)
testPics2	Image info object generated by kerasCreateDataset_2d
Learning_rate_l2	Minimum CLR assigned from tuneCLR
Learning_rate_h2	Maximum CLR assigned from tuneCLR
pathOut	path where the generated model(s) will be saved

Examples

runCLR(optimizer2 = optimizer_sgd(lr=lr_max, decay=0),
denseLoop2,
trainLabels2 = trainLabels,
epochs2 = 5,
batch_size2,
dropout2 = 0.2,
patience2 = 30,
validation_split2 = 0.2,
convolutionalLoop2 ,
NO_pooling2 = 1,
testPics2 = testPics,
Learning_rate_l2 = Learning_rate_l,
Learning_rate_h2 = Learning_rate_h,
pathOut = "keras/3categories/bayesmodels/"
)

kristianHoden/smartPARE documentation built on July 3, 2021, 7:10 p.m.