build_dnn: Build a Deep Neural Network for Survival Analysis

In survdnn: Deep Neural Networks for Survival Analysis with R 'torch'

Build a Deep Neural Network for Survival Analysis

Description

Constructs a multilayer perceptron (MLP) with optional batch normalization and dropout. Used internally by [survdnn()] to define the model architecture.

Usage

build_dnn(
  input_dim,
  hidden,
  activation = "relu",
  output_dim = 1L,
  dropout = 0.3,
  batch_norm = TRUE
)

Arguments

input_dim	Integer. Number of input features.
hidden	Integer vector. Sizes of the hidden layers (e.g., c(32, 16)).
activation	Character. Name of the activation function to use in each layer. Supported options: '"relu"', '"leaky_relu"', '"tanh"', '"sigmoid"', '"gelu"', '"elu"', '"softplus"'.
output_dim	Integer. Output layer dimension (default = 1).
dropout	Numeric between 0 and 1. Dropout rate after each hidden layer (default = 0.3). Set to 0 to disable dropout.
batch_norm	Logical; whether to add 'nn_batch_norm1d()' after each hidden linear layer (default = TRUE).

Value

A 'nn_sequential' object representing the network.

survdnn documentation built on Jan. 8, 2026, 9:07 a.m.