To implement RNN in an R environment, we can use the rnn package available through CRAN. This package is widely used to implement an RNN. A brief description of the rnn package, extracted from the official documentation, is shown in the following table:
| rnn: Recurrent Neural Network |
| Description: |
| Implementation of an RNN in R |
| Details: |
| Package: rnn Type: Package Version: 0.8.0 Date: 2016-09-11 License: GPL-3 |
| Authors: |
|
Bastiaan Quast Dimitri Fichou |
The main functions used from the rnn package are shown in this table:
|
predict_rnn |
Predicts the output of an RNN model: predict_rnn(model, X, hidden = FALSE, real_output = T, ...) |
|
run.rnn_demo |
A function to launch the rnn_demo app: run.rnn_demo(port ... |
