build_dnn

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

internal

Provides deep learning models for right-censored survival data using the 'torch' backend.
Supports multiple loss functions, including Cox partial likelihood, L2-penalized Cox, time-dependent Cox,
and accelerated failure time (AFT) loss. Offers a formula-based interface, built-in support for cross-validation,
hyperparameter tuning, survival curve plotting, and evaluation metrics such as the C-index, Brier score,
and integrated Brier score. For methodological details, see Kvamme et al. (2019) <https://www.jmlr.org/papers/v20/18-424.html>.

Imad El Badisy

survdnn

Deep Neural Networks for Survival Analysis Using 'torch'

Imad EL BADISY

build_dnn function

<dl><dt>input_dim</dt>
<dd>Integer. Number of input features.</dd>
<dt>hidden</dt>
<dd>Integer vector. Sizes of the hidden layers (e.g., c(32, 16)).</dd>
<dt>activation</dt>
<dd>Character. Name of the activation function to use in each layer.
Supported options: `"relu"`, `"leaky_relu"`, `"tanh"`, `"sigmoid"`, `"gelu"`, `"elu"`, `"softplus"`.</dd>
<dt>output_dim</dt>
<dd>Integer. Output layer dimension (default = 1).</dd></dl>

Arguments

Build a Deep Neural Network for Survival Analysis — build_dnn

<dl>

<dt>input_dim</dt>
<dd>Integer. Number of input features.</dd>


<dt>hidden</dt>
<dd>Integer vector. Sizes of the hidden layers (e.g., c(32, 16)).</dd>


<dt>activation</dt>
<dd>Character. Name of the activation function to use in each layer.
Supported options: `"relu"`, `"leaky_relu"`, `"tanh"`, `"sigmoid"`, `"gelu"`, `"elu"`, `"softplus"`.</dd>


<dt>output_dim</dt>
<dd>Integer. Output layer dimension (default = 1).</dd>

</dl>

build_dnn: Build a Deep Neural Network for Survival Analysis

Description

Usage

Value

Arguments

Examples