FTMixer: Frequency and Time Domain Representations Fusion for Time Series Modeling

TL;DR

FTMixer effectively combines frequency and time domain representations using novel modules and DCT, leading to improved long-term time series forecasting performance and efficiency.

Contribution

The paper introduces FTMixer, a novel model that fuses frequency and time domain features using FCC and WFC modules with DCT, enhancing global and local dependency capture.

Findings

Outperforms existing methods in forecasting accuracy

Demonstrates computational efficiency across datasets

Effectively captures both local and global dependencies

Abstract

Time series data can be represented in both the time and frequency domains, with the time domain emphasizing local dependencies and the frequency domain highlighting global dependencies. To harness the strengths of both domains in capturing local and global dependencies, we propose the Frequency and Time Domain Mixer (FTMixer). To exploit the global characteristics of the frequency domain, we introduce the Frequency Channel Convolution (FCC) module, designed to capture global inter-series dependencies. Inspired by the windowing concept in frequency domain transformations, we present the Windowing Frequency Convolution (WFC) module to capture local dependencies. The WFC module first applies frequency transformation within each window, followed by convolution across windows. Furthermore, to better capture these local dependencies, we employ channel-independent scheme to mix the time domain and frequency domain patches. Notably, FTMixer employs the Discrete Cosine Transformation (DCT) with real numbers instead of the complex-number-based Discrete Fourier Transformation (DFT), enabling direct utilization of modern deep learning operators in the frequency domain. Extensive experimental results across seven real-world long-term time series datasets demonstrate the superiority of FTMixer, in terms of both forecasting performance and computational efficiency.