Recurrent Neural Network (RNN) and its variations such as Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU), have become standard building blocks for learning online data of sequential nature in many research areas, including natural language processing and speech data analysis. In this paper, we present a new methodology to significantly reduce the number of parameters in RNNs while maintaining performance that is comparable or even better than classical RNNs. The new proposal, referred to as Restricted Recurrent Neural Network (RRNN), restricts the weight matrices corresponding to the input data and hidden states at each time step to share a large proportion of parameters. The new architecture can be regarded as a compression of its classical counterpart, but it does not require pre-training or sophisticated parameter fine-tuning, both of which are major issues in most existing compression techniques. Experiments on natural language modeling show that compared with its classical counterpart, the restricted recurrent architecture generally produces comparable results at about 50% compression rate. In particular, the Restricted LSTM can outperform classical RNN with even less number of parameters.
|Original language||English (US)|
|Title of host publication||Proceedings - 2019 IEEE International Conference on Big Data, Big Data 2019|
|Editors||Chaitanya Baru, Jun Huan, Latifur Khan, Xiaohua Tony Hu, Ronay Ak, Yuanyuan Tian, Roger Barga, Carlo Zaniolo, Kisung Lee, Yanfang Fanny Ye|
|Publisher||Institute of Electrical and Electronics Engineers Inc.|
|Number of pages||8|
|State||Published - Dec 2019|
|Event||2019 IEEE International Conference on Big Data, Big Data 2019 - Los Angeles, United States|
Duration: Dec 9 2019 → Dec 12 2019
|Name||Proceedings - 2019 IEEE International Conference on Big Data, Big Data 2019|
|Conference||2019 IEEE International Conference on Big Data, Big Data 2019|
|Period||12/9/19 → 12/12/19|
Bibliographical noteFunding Information:
This work was supported in part by Office of Naval Research Grant No. N00014-18-1-2244.
This work was supported in part by Office of Naval Research Grant No. N00014-18-1-2244. We provide our implementation at https://github.com/ dem123456789/Restricted-Recurrent-Neural-Networks. 978-1-7281-0858-2/19/$31.00 © 2019 IEEE
© 2019 IEEE.
- Gated Recurrent Unit
- Long Short-Term Memory
- Model Compression
- Parameter Sharing
- Recurrent Neural Networks