Back Propagation Through Time for Recurrent Neural Networks

Neural networks
Mathematics
Author
Affiliation

Ramkumar

Research Scholar, IIST

Published

April 27, 2025

1 Preface

Recurrent Neural Networks are the type of ANNs that are used for processing sequential data such as a string of characters and time-series data. These RNNs can be trained to perform inference on each timestep data (eg. word completion) or to infer the output at the end of sequence (eg. movie review).

Following were discussed in this writeup

  1. Overview on RNNs
  2. Mathematical background of RNNs
  3. Back Propagation Through Time
  4. Training algorithm of RNNs

The writeup can be opened with this link.