Deep Autoregressive Models with Spectral Attention
This work addresses time series forecasting for domains like real-world applications, but it appears incremental as it builds on existing architectures with a novel attention mechanism.
The authors tackled time series forecasting by proposing a deep autoregressive model with a Spectral Attention module that merges global and local frequency domain information to identify trends and seasonality while filtering noise, resulting in improved forecasting accuracy compared to state-of-the-art approaches on multiple datasets.
Time series forecasting is an important problem across many domains, playing a crucial role in multiple real-world applications. In this paper, we propose a forecasting architecture that combines deep autoregressive models with a Spectral Attention (SA) module, which merges global and local frequency domain information in the model's embedded space. By characterizing in the spectral domain the embedding of the time series as occurrences of a random process, our method can identify global trends and seasonality patterns. Two spectral attention models, global and local to the time series, integrate this information within the forecast and perform spectral filtering to remove time series's noise. The proposed architecture has a number of useful properties: it can be effectively incorporated into well-know forecast architectures, requiring a low number of parameters and producing interpretable results that improve forecasting accuracy. We test the Spectral Attention Autoregressive Model (SAAM) on several well-know forecast datasets, consistently demonstrating that our model compares favorably to state-of-the-art approaches.