A Memory-Network Based Solution for Multivariate Time-Series Forecasting

TL;DR

This paper introduces MTNet, a memory-augmented deep learning model for multivariate time-series forecasting that captures complex patterns, long-term dependencies, and offers interpretability through an attention mechanism.

Contribution

The paper presents MTNet, a novel memory network architecture specifically designed for multivariate time-series forecasting, addressing long-term dependency and interpretability issues.

Findings

MTNet outperforms traditional models on multiple datasets.

The attention mechanism enhances model interpretability.

MTNet effectively captures long-term dependencies.

Abstract

Multivariate time series forecasting is extensively studied throughout the years with ubiquitous applications in areas such as finance, traffic, environment, etc. Still, concerns have been raised on traditional methods for incapable of modeling complex patterns or dependencies lying in real word data. To address such concerns, various deep learning models, mainly Recurrent Neural Network (RNN) based methods, are proposed. Nevertheless, capturing extremely long-term patterns while effectively incorporating information from other variables remains a challenge for time-series forecasting. Furthermore, lack-of-explainability remains one serious drawback for deep neural network models. Inspired by Memory Network proposed for solving the question-answering task, we propose a deep learning based model named Memory Time-series network (MTNet) for time series forecasting. MTNet consists of a large memory component, three separate encoders, and an autoregressive component to train jointly. Additionally, the attention mechanism designed enable MTNet to be highly interpretable. We can easily tell which part of the historic data is referenced the most.