Scribd
Upload
138 views22 pages

Long Short-Term Memory Networks PDF

Long Short-Term Memory (LSTM) is a type of recurrent neural network (RNN) that addresses the vanishing gradient problem and allows learning of long-term dependencies. An LSTM memory cell contains gates that regulate the flow of information into and out of the cell through multiplicative interactions. These gates allow the network to store and access information over long periods of time. LSTMs have achieved state-of-the-art results on tasks involving sequence modeling such as speech recognition, machine translation, and image captioning.

Uploaded by

Subham Singh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
138 views22 pages

Long Short-Term Memory Networks PDF

Long Short-Term Memory (LSTM) is a type of recurrent neural network (RNN) that addresses the vanishing gradient problem and allows learning of long-term dependencies. An LSTM memory cell contains gates that regulate the flow of information into and out of the cell through multiplicative interactions. These gates allow the network to store and access information over long periods of time. LSTMs have achieved state-of-the-art results on tasks involving sequence modeling such as speech recognition, machine translation, and image captioning.

Uploaded by

Subham Singh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 22

Long Short-Term Memory

Akshay Sood
Introduction
● Feedforward neural networks
Recurrent Neural Networks (RNNs)
● Networks with feedback loops (recurrent edges)
● Output at current time step depends on current input as well as previous state (via recurrent edges)

UNFOLD
IN TIME
Training RNNs
● Backpropagation Through Time (BPTT)
○ Regular (feedforward) backprop applied to RNN unfolded in time
○ Truncated BPTT approximation
Training RNNs
● Problem: can’t capture long-term dependencies due to vanishing/exploding
gradients during backpropagation

UNFOLD
Long Short-Term Memory networks (LSTMs)
● A type of RNN architecture that addresses the vanishing/exploding gradient problem
and allows learning of long-term dependencies

● Recently risen to prominence with state-of-the-art performance in speech recognition,


language modeling, translation, image captioning
LSTMs
Central Idea: A memory cell (interchangeably block) which can maintain its state
over time, consisting of an explicit memory (aka the cell state vector) and gating
units which regulate the information flow into and out of the memory.

MEMORY

LSTM Memory Cell


LSTM Memory Cell
Gate (sigmoid layer
followed by pointwise
multiplication)

Simplified schematic
for reference
Cell state vector
● Represents the memory of the LSTM
● Undergoes changes via forgetting of old memory (forget gate) and addition of new
memory (input gate)

Cell state vector


Gates

● Gate: sigmoid neural net layer followed by pointwise multiplication operator

● Gates control the flow of information to/from the memory

● Gates are controlled by a concatenation of the output from the previous time step and
the current input and optionally the cell state vector.
Forget Gate
● Controls what information to throw away from memory
Input Gate
● Controls what new information is added to cell state from current input
Memory Update
● The cell state vector aggregates the two components (old memory via the
forget gate and new memory via the input gate)
Output Gate
● Conditionally decides what to output from the memory
LSTM Memory Cell Summary
LSTM Training
● Backpropagation Through Time (BPTT) most common
● What weights are learned?
○ Gates (input/output/forget)
○ Input tanh layer
● Outputs depend on the task:
○ Single output prediction for the whole sequence (e.g. below)
○ One output at each time step (sequence labeling)
Deep LSTMs

● Deep LSTMs can be created by stacking multiple LSTM


layers vertically, with the output sequence of one layer
forming the input sequence of the next (in addition to
recurrent connections within the same layer)

● Increases the number of parameters - but given sufficient


data, performs significantly better than single-layer LSTMs
(Graves et al. 2013)

● Dropout usually applied only to non-recurrent edges,


including between layers
Bidirectional RNNs
● Data processed in both directions processed with two separate hidden layers, which are then fed
forward into the same output layer

● Bidirectional RNNs can better exploit context in both directions, for e.g. bidirectional LSTMs perform
better than unidirectional ones in speech recognition (Graves et al. 2013)
LSTMs for Machine Translation (Sutskever et al. 2014)

● Encoder and decoder LSTMs


Demos
● Handwriting generation demo:
http://www.cs.toronto.edu/~graves/handwriting.html
● Music composition:
http://www.hexahedria.com/2015/08/03/composing-music-with-recurrent-neur
al-networks/
● Image captioning and other stuff:
http://karpathy.github.io/2015/05/21/rnn-effectiveness/
References
● Hochreiter, Sepp, and Jürgen Schmidhuber. "Long short-term memory." Neural computation 9.8 (1997): 1735-1780.
(The original paper on LSTMs; the forget gate was added later)
● http://colah.github.io/posts/2015-08-Understanding-LSTMs/ (A great blog post introducing LSTMs)
● Lipton, Zachary C., John Berkowitz, and Charles Elkan. "A critical review of recurrent neural networks for sequence
learning." (A nice review of RNNs, including LSTMs, bidirectional RNNs and state-of-the-art applications)
● https://deeplearning4j.org/lstm (Another nice introduction to recurrent networks and LSTMs, with code examples -
Deeplearning4j is a deep learning platform for Java and Scala)
● Sutskever, I., Vinyals, O., & Le, Q. (2014). Sequence to sequence learning with neural networks. Advances in Neural
Information. (A paper that proposes two LSTMs (one for encoding, one for decoding) for machine translation)
● Graves, A., Mohamed, A., & Hinton, G. (2013). Speech recognition with deep recurrent neural networks. Acoustics,
Speech and Signal. (A paper that proposes deep bidirectional LSTMs for speech recognition)
● Karpathy, Andrej, and Li Fei-Fei. "Deep visual-semantic alignments for generating image descriptions." Proceedings
of the IEEE Conference on Computer Vision and Pattern Recognition. 2015. (Paper introducing image captioning
using ConvNet + LSTM)
● https://medium.com/@shiyan/understanding-lstm-and-its-diagrams-37e2f46f1714 (Neat LSTM explanation diagrams)
● http://deeplearning.net/tutorial/lstm.html (Tutorial applying LSTM to sentiment analysis)
● https://xkcd.com/1093/
Other useful links
● http://deeplearning.net/tutorial/lstm.html
● https://github.com/zhongkaifu/RNNSharp
● http://blog.leanote.com/post/[email protected]/RNN-and-LSTM-List
● https://deeplearning4j.org/lstm
● https://apaszke.github.io/lstm-explained.html
● https://medium.com/@shiyan/understanding-lstm-and-its-diagrams-37e2f46f1
714

You might also like