Introduction to Deep Learning Using R: A Step-by-Step Guide to Learning and Implementing Deep Learning Models Using R

© Taweh Beysolow II 2017

Taweh Beysolow II, Introduction to Deep Learning Using R, https://doi.org/10.1007/978-1-4842-2734-3_6

6. Recurrent Neural Networks (RNNs)

Taweh Beysolow II¹ 

(1)San Francisco, California, USA

Recurrent neural networks (RNNs) are models that were created to tackle problems within the scope of pattern recognition and are fundamentally built on the same concepts with respect to feed-forward MLPs. The difference is that although MLPs by definition have multiple layers, RNNs do not and instead have a directed cycle through which the inputs are transformed into outputs. I’ll begin the chapter by covering several RNN models and end it with a practical application of RNNs.

Fully Recurrent Networks

Imagine that we have an input, ...