Corpus ID: 3276568

Noisin: Unbiased Regularization for Recurrent Neural Networks

@article{Dieng2018NoisinUR,
  title={Noisin: Unbiased Regularization for Recurrent Neural Networks},
  author={Adji B. Dieng and R. Ranganath and Jaan Altosaar and D. Blei},
  journal={ArXiv},
  year={2018},
  volume={abs/1805.01500}
}

• Adji B. Dieng, R. Ranganath, +1 author D. Blei
• Published 2018
• Computer Science, Mathematics
• ArXiv

Recurrent neural networks (RNNs) are powerful models of sequential data. They have been successfully used in domains such as text and speech. However, RNNs are susceptible to overfitting; regularization is important. In this paper we develop Noisin, a new method for regularizing RNNs. Noisin injects random noise into the hidden states of the RNN and then maximizes the corresponding marginal likelihood of the data. We show how Noisin applies to any RNN and we study many different types of noise… Expand

View PDF on arXiv

Save to Library

Create Alert

Cite

Launch Research Feed