TLNets: Transformation Learning Networks for long-range time-series prediction

TL;DR

TLNets are a novel class of networks for long-range time-series prediction that leverage transformation mechanisms like Fourier Transform, SVD, and convolution to effectively fuse features across scales and improve forecasting performance.

Contribution

The paper introduces Transformation Learning Networks (TLNets) with four transformation-based modules, enhancing receptive fields and feature fusion for long-range time-series prediction.

Findings

Outperform baseline models on real-world datasets.

Effective fusion of global and local features.

Potential for improved long-range forecasting accuracy.

Abstract

Time series prediction is a prevalent issue across various disciplines, such as meteorology, traffic surveillance, investment, and energy production and consumption. Many statistical and machine-learning strategies have been developed to tackle this problem. However, these approaches either lack explainability or exhibit less satisfactory performance when the prediction horizon increases. To this end, we propose a novel plan for the designing of networks' architecture based on transformations, possessing the potential to achieve an enhanced receptive field in learning which brings benefits to fuse features across scales. In this context, we introduce four different transformation mechanisms as bases to construct the learning model including Fourier Transform (FT), Singular Value Decomposition (SVD), matrix multiplication and Conv block. Hence, we develop four learning models based on the above building blocks, namely, FT-Matrix, FT-SVD, FT-Conv, and Conv-SVD. Note that the FT and SVD blocks are capable of learning global information, while the Conv blocks focus on learning local information. The matrix block is sparsely designed to learn both global and local information simultaneously. The above Transformation Learning Networks (TLNets) have been extensively tested and compared with multiple baseline models based on several real-world datasets and showed clear potential in long-range time-series forecasting.