Autoregressive Convolutional Neural Networks for Asynchronous Time Series
This work addresses time series prediction for applications like finance and energy consumption, but it appears incremental as it builds on existing autoregressive and neural network methods.
The authors tackled regression of multivariate asynchronous time series by proposing the Significance-Offset Convolutional Neural Network, achieving promising results compared to convolutional and recurrent neural networks on datasets including a hedge fund proprietary dataset with over 2 million quotes.
We propose Significance-Offset Convolutional Neural Network, a deep convolutional network architecture for regression of multivariate asynchronous time series. The model is inspired by standard autoregressive (AR) models and gating mechanisms used in recurrent neural networks. It involves an AR-like weighting system, where the final predictor is obtained as a weighted sum of adjusted regressors, while the weights are datadependent functions learnt through a convolutional network. The architecture was designed for applications on asynchronous time series and is evaluated on such datasets: a hedge fund proprietary dataset of over 2 million quotes for a credit derivative index, an artificially generated noisy autoregressive series and UCI household electricity consumption dataset. The proposed architecture achieves promising results as compared to convolutional and recurrent neural networks.