Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) Network
By A. Sherstinsky. Physica A: Statistical Mechanics and its Applications, 2018
5,102 citations recorded by the source corpus.
Abstract Because of their effectiveness in broad practical applications, LSTM networks have received a wealth of coverage in scientific journals, technical blogs, and implementation guides. However, in most articles, the inference formulas for the LSTM network and its parent, RNN, are stated axiomatically, while the training formulas are omitted altogether. In addition, the technique of “unrolling” an RNN is routinely presented without justification throughout the literature. The goal of this tutorial is to explain the essential RNN and LSTM fundamentals in a single document. Drawing from concepts in Signal Processing, we formally derive the canonical RNN formulation from differential equations. We then propose and prove a precise statement, which yields the RNN unrolling technique. We also review the difficulties with training the standard RNN and address them by transforming the RNN into the “Vanilla LSTM” 1 network through a series of logical arguments. We provide all equations pertaining to the LSTM system together with detailed descriptions of its constituent entities. Albeit unconventional, our choice of notation and the method for presenting the LSTM system emphasizes ease of understanding. As part of the analysis, we identify new opportunities to enrich the LSTM system and incorporate these extensions into the Vanilla LSTM network, producing the most general LSTM variant to date. The target reader has already been exposed to RNNs and LSTM networks through numerous available resources and is open to an alternative pedagogical approach. A Machine Learning practitioner seeking guidance for implementing our new augmented LSTM model in software for experimentation and research will find the insights and derivations in this treatise valuable as well.
Subfields: Artificial Intelligence, Statistical and Nonlinear Physics, Signal Processing
View Paper: https://www.semanticscholar.org/paper/2113af2c71526b211612586daa185d326712cb9c
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Jens Tuyls
- Joaquim R. Martins
- Ji-chao Li
- Dylan Slack
- Gavin Kerrigan
- Lizhi Wang
- Sotirios V. Archontoulis
- Xiaosong Du
In Statistical and Nonlinear Physics
In Signal Processing
Citations in the corpus
- Using labeled data to evaluate change detectors in a multivariate streaming environment (2009, neutral): “Furthermore, since there are many small ice blocks in the river ice images, a learnable upsampling strategy DUpsampling is utilized to replace the commonly used bilinear interpolation to increase the segmentation accuracy, especially for small ice blocks.”
- Differentially Private Language Models Benefit from Public Pre-training (2020, background): “Many high-quality language models rely on some form of recurrent neural architecture, such as RNNs or LSTMs (Sherstinsky, 2018; Hochre-iter and Schmidhuber, 1997).”
- A CNN-RNN Framework for Crop Yield Prediction (2019, methodology): “RNNs are used for tasks involving sequential data to capture their time dependencies (LeCun et al., 2015; Sherstinsky, 2018).”
- Machine Learning in Aerodynamic Shape Optimization (2022, background): “Recurrent Neural Networks RNN gets its name from a feedback loop that points a hidden neuron back to itself, constituting a recurrent structure [257, 258].”