nn_model

This function defines a generative neural network model for a certain architecture
and adds noise to the input layer.

internal

Distributional instrumental variable (DIV) model
for estimation of the interventional distribution of the outcome Y under a
do-intervention on the treatment X. Instruments, predictors and targets can
be univariate or multivariate. Functionality includes estimation of the (conditional)
interventional mean and quantiles, as well as sampling from the fitted (conditional)
interventional distribution.

Anastasiia Holovchak

DistributionIV

Distributional Instrumental Variable (DIV) Model

Sorawit Saengkyongam

Nicolai Meinshausen

Xinwei Shen

nn_model function

<dl><dt>in_dim</dt>
<dd>Integer. Input dimension.</dd>
<dt>noise_dim</dt>
<dd>Integer. Dimension of the noise to inject.</dd>
<dt>hidden_dim</dt>
<dd>Integer. Number of neurons in the hidden layers (default: 100).</dd>
<dt>out_dim</dt>
<dd>Integer. Output dimension.</dd>
<dt>num_layer</dt>
<dd>Integer. Number of layers in the model (default: 3).</dd></dl>

Arguments

Define a Stochastic Generative Neural Network Model with Noise at Input Layer — nn_model

<dl>

<dt>in_dim</dt>
<dd>Integer. Input dimension.</dd>


<dt>noise_dim</dt>
<dd>Integer. Dimension of the noise to inject.</dd>


<dt>hidden_dim</dt>
<dd>Integer. Number of neurons in the hidden layers (default: 100).</dd>


<dt>out_dim</dt>
<dd>Integer. Output dimension.</dd>


<dt>num_layer</dt>
<dd>Integer. Number of layers in the model (default: 3).</dd>

</dl>

nn_model: Define a Stochastic Generative Neural Network Model with Noise at Input Layer

Description

Usage

Value

Arguments