dnnControl {dnn}

Auxiliary function for dnnFit dnnFit

Description

dnnControl is an auxiliary function for dnnFit. Typically only used internally by the dnn package, may be used to construct a control argument for the deep learning neural network model to specify parameters such as a loss function.

Usage

  dnnControl(loss = c("mse", "cox", "bin", "log", "mae"), epochs = 300, 
	     batch_size = 64, verbose = 0, lr_rate = 0.0001,  
	     alpha = 0.5, lambda = 1.0, epsilon = 0.01, max.iter = 100, 
	     censor.group = NULL, weights = NULL)

Arguments

loss

loss function for the neural network model, "mse" for mean square error (guassian glm model), "mae" for mean absolute error, "cox" for the Cox partial likelihood (proportional hazards model), "bin" for cross-entropy (binomial glm model), "log" for log-linear (poisson glm model).

epochs

number of deep learning epochs, default is 30.

batch_size

batch size, default is 64. 'NaN' may be generated if batch size is too small and there is not event in a batch.

lr_rate

learning rate, default is 0.0001.

weights

an optional vector of 'prior weights' to be used in the fitting process. Should be NULL or a numeric vector, default is NULL.

alpha

alpha decay rate for momentum gradient descent, default is 0.5.

lambda

regularization term for dnn weighting parameters, 0.5*lambda*W*W), default is 1.0.

verbose

verbose = 1 for print out verbose during the model fit, 0 for not print.

epsilon

epsilon for convergence check, default is epsilon = 0.01.

max.iter

number of maximum iteration, default is max.iter = 100. This is used in the deepAFT function

censor.group

a vector for censoring groups. A KM curve for censoring will be fit for each group. If a matrix is provided, then a Cox model will be used to predict the censoring probability. Used only in the deepAFT function.

Details

dnnControl is used in model fitting of "dnnFit". Additional loss functions will be added to the library in the future.

Value

This function checks the internal consistency and returns a list of values as input to control model fitting of "dnnFit".

Note

For right censored survival time only

Author(s)

Chen, B. E.

References

Norman, P. and Chen, B. E. (2023). DeepAFAT: A nonparametric accelerated failure time model with artificial neural network. Manuscript to be submitted.

See Also

deepAFT, deepGLM, deepSurv, dnnFit

Examples

## Example for dnnControl
##
# model = dNNmodel()

  control = dnnControl(loss='mse')
  
# can also be used in   
# fit = dnnFit(y ~ x, model, control) 
# print(fit)