2015/04/07 by Mathias Berglund, Tapani Raiko, Berglund, Mathias +9 · 1 citation
Computer Science · #FOS: Computer and information sciences #Machine Learning (cs.LG) #Music and Audio Processing #Neural Networks and Applications #Neural and Evolutionary Computing (cs.NE) #Time Series Analysis and Forecasting
paper · pdf · doi:10.48550/arxiv.1504.01575
openalex publication_date 2015/04/07 · openalex created_date 2016/06/24 · openalex updated_date 2026/07/28
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 between reconstruction and probabilistic modeling, which makes the interpretation possible. As far as we know, neither GSN or NADE have been studied in the context of time series before. As an example of an unsupervised task, we study the problem of filling in gaps in high-dimensional time series with complex dynamics. Although unidirectional RNNs have recently been trained successfully to model such time series, inference in the negative time direction is non-trivial. We propose two probabilistic interpretations of bidirectional RNNs that can be used to reconstruct missing gaps efficiently. Our experiments on text data show that both proposed methods are much more accurate than unidirectional reconstructions, although a bit less accurate than a computationally complex bidirectional Bayesian inference on the unidirectional RNN. We also provide results on music data for which the Bayesian inference is computationally infeasible, demonstrating the scalability of the proposed methods.