We propose here an extended attention model for sequence-to-sequence recurrent neural networks (RNNs) designed to capture (pseudo-)periods in time series. This extended attention model can be deployed on top of any RNN and is shown to yield state-of-the-art performance for time series forecasting on several univariate and multivariate time series.
@article{arxiv.1703.10089,
title = {Position-based Content Attention for Time Series Forecasting with Sequence-to-sequence RNNs},
author = {Yagmur G. Cinar and Hamid Mirisaee and Parantapa Goswami and Eric Gaussier and Ali Ait-Bachir and Vadim Strijov},
journal= {arXiv preprint arXiv:1703.10089},
year = {2017}
}