Bidirectional Recurrent Neural Networks as Generative Models - Reconstructing Gaps in Time Series

CoRR, 2015.

Cited by: 10|Bibtex|Views50|Links
EI

Abstract:

Bidirectional recurrent neural networks (RNN) are trained to predict both in the positive and negative time directions simultaneously. They have not been used commonly in unsupervised tasks, because a probabilistic interpretation of the model has been difficult. Recently, two different frameworks, GSN and NADE, provide a connection betw...More

Code:

Data:

Your rating :
0

 

Tags
Comments