cito1.1 package

Building and Training Neural Networks

ALE

Accumulated Local Effect Plot (ALE)

analyze_training

Visualize training of Neural Network

avgPool

Average pooling layer

cito

'cito': Building and training neural networks

cnn

CNN

coef.citocnn

Returns list of parameters the neural network model currently has in u...

coef.citodnn

Returns list of parameters the neural network model currently has in u...

conditionalEffects

Calculate average conditional effects

config_lr_scheduler

Creation of customized learning rate scheduler objects

config_optimizer

Creation of customized optimizer objects

config_tuning

Config hyperparameter tuning

continue_training

Continues training of a model generated with dnn or cnn for additi...

conv

Convolutional layer

create_architecture

CNN architecture

dnn

DNN

e

Embeddings

findReTrmClasses

list of specials -- taken from enum.R

linear

Linear layer

maxPool

Maximum pooling layer

PDP

Partial Dependence Plot (PDP)

plot.citoarchitecture

Plot the CNN architecture

plot.citocnn

Plot the CNN architecture

plot.citodnn

Creates graph plot which gives an overview of the network architecture...

predict.citocnn

Predict from a fitted cnn model

predict.citodnn

Predict from a fitted dnn model

print.avgPool

Print pooling layer

print.citoarchitecture

Print class citoarchitecture

print.citocnn

Print class citocnn

print.citodnn

Print class citodnn

print.conditionalEffects

Print average conditional effects

print.conv

Print conv layer

print.linear

Print linear layer

print.maxPool

Print pooling layer

print.summary.citodnn

Print method for class summary.citodnn

print.transfer

Print transfer model

residuals.citodnn

Extract Model Residuals

simulate_shapes

Data Simulation for CNN

summary.citocnn

Summary citocnn

summary.citodnn

Summarize Neural Network of class citodnn

sumTerms

combine a list of formula terms as a sum

transfer

Transfer learning

tune

Tune hyperparameter

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.

  • Maintainer: Maximilian Pichler
  • License: GPL (>= 3)
  • Last published: 2024-03-18