PAPERLong Short-Term Memory (Optional- original paper)

Covers: theory of Long short-term memory (LSTM)
Estimated time needed: 30 minutes
Questions this item adddesses:
  • Where does this model come from?
How to use this item?

Read the whole article

URL: https://www.bioinf.jku.at/publications/older/2604.pdf
Author(s) / creator(s) / reference(s)
Hochreiter
Shortlist
publicShare

LSTM

Amir PariziTotal time needed: ~3 hours
Learning Objectives
Long short-term memory (LSTM) is an artificial recurrent neural network (RNN) architecture used in the field of deep learning
Potential Use Cases
Robot control, Time series prediction,Speech recognition,Rhythm learning, Music composition, Grammar learning, Handwriting recognition, Human action recognition.
Target Audience
ADVANCED
Go through the following annotated items in order:
VIDEO 1. Introduction to RNN
  • What is RNN?
30 minutes
LIBRARY 2. RNN from scratch
  • How to implement RNN in python from scratch?
20 minutes
VIDEO 3. Long Short Term Memory (LSTM) - Introduction
  • What is LSTM and what is the math behind this model?
20 minutes
ARTICLE 4. Understanding LSTM
  • What is LSTM and how different compartment work in this model?
25 minutes
LIBRARY 5. Implementing LSTM in python
  • How to implement LSTM in python?
30 minutes
PAPER 6. Long Short-Term Memory (Optional- original paper)
  • Where does this model come from?
30 minutes

Concepts Convered