backprop_surv: Backprop Surv (Internal)
In SkadiEye/deepTL: Deep Treatment Learning

backprop_surv

R Documentation

Backprop Surv (Internal)

Description

Backprop Surv (Internal)

Usage

backprop_surv(
  n_hidden,
  w_ini,
  load_param,
  weight,
  bias,
  x,
  y,
  w,
  valid,
  x_valid,
  y_valid,
  w_valid,
  activ,
  n_epoch,
  n_batch,
  model_type,
  learning_rate,
  l1_reg,
  l2_reg,
  early_stop,
  early_stop_det,
  learning_rate_adaptive,
  rho,
  epsilon,
  beta1,
  beta2,
  loss_f
)

Arguments

`n_hidden`	Hidden layer numbers.
`w_ini`	Initial weights.
`load_param`	Whether to load parameters.
`weight`	Weights.
`bias`	Biases.
`x`	Training.
`y`	Training.
`w`	Training.
`valid`	Whether to use a validation set.
`x_valid`	Validation.
`y_valid`	Validation.
`w_valid`	Validation.
`activ`	Activation function.
`n_epoch`	Number of epochs.
`n_batch`	Batch size.
`model_type`	Model type.
`learning_rate`	(Initial) learning rate.
`l1_reg`	L1-penalty.
`l2_reg`	L2-penalty.
`early_stop`	Whether to early stop.
`early_stop_det`	Number of epochs to determine early-stop.
`learning_rate_adaptive`	Adaptive learning rate adjustment method.
`rho`	Parameter.
`epsilon`	Parameter.
`beta1`	Parameter.
`beta2`	Parameter.
`loss_f`	Loss function.