Seasonal Encoder-Decoder Architecture for Forecasting
This work addresses forecasting challenges in time series analysis, offering a domain-specific solution that appears incremental by building on encoder-decoder and seasonal auto-regressive models.
The paper tackles time series forecasting by proposing a novel RNN architecture that captures seasonal correlations and enables accurate multi-step predictions, demonstrating its utility through extensive experiments in single and multiple sequence scenarios.
Deep learning (DL) in general and Recurrent neural networks (RNNs) in particular have seen high success levels in sequence based applications. This paper pertains to RNNs for time series modelling and forecasting. We propose a novel RNN architecture capturing (stochastic) seasonal correlations intelligently while capable of accurate multi-step forecasting. It is motivated from the well-known encoder-decoder (ED) architecture and multiplicative seasonal auto-regressive model. It incorporates multi-step (multi-target) learning even in the presence (or absence) of exogenous inputs. It can be employed on single or multiple sequence data. For the multiple sequence case, we also propose a novel greedy recursive procedure to build (one or more) predictive models across sequences when per-sequence data is less. We demonstrate via extensive experiments the utility of our proposed architecture both in single sequence and multiple sequence scenarios.