cito {cito}

R Documentation

'cito': Building and training neural networks

Description

Building and training custom neural networks in the typical R syntax. The 'torch' package is used for numerical calculations, which allows for training on CPU as well as on a graphics card. The main function is dnn which trains a custom deep neural network.

Installation

in order to install cito please follow these steps:

install.packges("cito")

library(torch)

install_torch(reinstall = TRUE)

library(cito)

cito functions

• dnn: train deep neural network

• continue_training: continues training of an existing cito dnn model for additional epochs

• PDP: plot the partial dependency plot for a specific feature

• ALE: plot the accumulated local effect plot for a specific feature

Examples

vignette("cito", package="cito")

---