Multi-scale Transformer Pyramid Networks for Multivariate Time Series Forecasting
This work addresses a limitation in transformer-based forecasting for multivariate time series, offering improved modeling of multiple unconstrained scales to handle patterns like hourly and daily seasonalities.
The paper tackles the problem of capturing diverse seasonalities in multivariate time series forecasting by introducing a Multi-scale Transformer Pyramid Network (MTPNet) with dimension invariant embedding, resulting in outperforming recent state-of-the-art methods on nine benchmark datasets.
Multivariate Time Series (MTS) forecasting involves modeling temporal dependencies within historical records. Transformers have demonstrated remarkable performance in MTS forecasting due to their capability to capture long-term dependencies. However, prior work has been confined to modeling temporal dependencies at either a fixed scale or multiple scales that exponentially increase (most with base 2). This limitation hinders their effectiveness in capturing diverse seasonalities, such as hourly and daily patterns. In this paper, we introduce a dimension invariant embedding technique that captures short-term temporal dependencies and projects MTS data into a higher-dimensional space, while preserving the dimensions of time steps and variables in MTS data. Furthermore, we present a novel Multi-scale Transformer Pyramid Network (MTPNet), specifically designed to effectively capture temporal dependencies at multiple unconstrained scales. The predictions are inferred from multi-scale latent representations obtained from transformers at various scales. Extensive experiments on nine benchmark datasets demonstrate that the proposed MTPNet outperforms recent state-of-the-art methods.