Clockwork RNN

swMATH ID: 36448
Software Authors: Jan Koutník, Klaus Greff, Faustino Gomez, Jürgen Schmidhuber
Description: A Clockwork RNN. Sequence prediction and classification are ubiquitous and challenging problems in machine learning that can require identifying complex dependencies between temporally distant inputs. Recurrent Neural Networks (RNNs) have the ability, in theory, to cope with these temporal dependencies by virtue of the short-term memory implemented by their recurrent (feedback) connections. However, in practice they are difficult to train successfully when the long-term memory is required. This paper introduces a simple, yet powerful modification to the standard RNN architecture, the Clockwork RNN (CW-RNN), in which the hidden layer is partitioned into separate modules, each processing inputs at its own temporal granularity, making computations only at its prescribed clock rate. Rather than making the standard RNN models more complex, CW-RNN reduces the number of RNN parameters, improves the performance significantly in the tasks tested, and speeds up the network evaluation. The network is demonstrated in preliminary experiments involving two tasks: audio signal generation and TIMIT spoken word classification, where it outperforms both RNN and LSTM networks.
Homepage: https://arxiv.org/abs/1402.3511
Source Code:  https://github.com/flandolfi/clockwork-rnn
Keywords: arXiv_cs.NE; Machine Learning; arXiv_cs.LG; Clockwork RNN; Recurrent Neural Networks; RNNs; CW-RNN
Related Software: Evolino; Zoneout; Adam; ImageNet; RMSprop; LSTM; DARPA TIMIT; IMDB; UCF101; word2vec; Skip RNN; MNIST; BinaryConnect; Theano; Penn Treebank; SenticNet; MS-COCO; Flickr30K; AlexNet; SpiNNaker
Cited in: 1 Document

Standard Articles

1 Publication describing the Software Year
A Clockwork RNN arXiv
Jan Koutník, Klaus Greff, Faustino Gomez, Jürgen Schmidhuber

Cited in 1 Serial

1 Neural Computation

Citations by Year