Can be used for categorical variables, a more efficient alternative to one-hot encoding

The 'cito' package provides a user-friendly interface for training and interpreting deep neural networks (DNN). 'cito' simplifies the fitting of DNNs by supporting the familiar formula syntax, hyperparameter tuning under cross-validation, and helps to detect and handle convergence problems.  DNNs can be trained on CPU, GPU and MacOS GPUs. In addition, 'cito' has many downstream functionalities such as various explainable AI (xAI) metrics (e.g. variable importance, partial dependence plots, accumulated local effect plots, and effect estimates) to interpret trained DNNs. 'cito' optionally provides confidence intervals (and p-values) for all xAI metrics and predictions. At the same time, 'cito' is computationally efficient because it is based on the deep learning framework 'torch'. The 'torch' package is native to R, so no Python installation or other API is required for this package.

Maximilian Pichler

cito

Building and Training Neural Networks

Christian Amesöder

Florian Hartig

Armin Schenk

e function

<dl><dt>dim</dt>
<dd>integer, embedding dimension</dd>
<dt>weights</dt>
<dd>matrix, to use custom embedding matrices</dd>
<dt>train</dt>
<dd>logical, should the embeddings be trained or not</dd>
<dt>lambda</dt>
<dd>regularization strength on the embeddings</dd>
<dt>alpha</dt>
<dd>mix between L1 and L2 regularization</dd></dl>

Arguments

Embeddings — e

<dl>

<dt>dim</dt>
<dd>integer, embedding dimension</dd>


<dt>weights</dt>
<dd>matrix, to use custom embedding matrices</dd>


<dt>train</dt>
<dd>logical, should the embeddings be trained or not</dd>


<dt>lambda</dt>
<dd>regularization strength on the embeddings</dd>


<dt>alpha</dt>
<dd>mix between L1 and L2 regularization</dd>

</dl>

e: Embeddings

Description

Usage

Arguments