Interpretable Multivariate Time Series Forecasting Using Neural Fourier Transform

TL;DR

This paper introduces the Neural Fourier Transform, a novel model combining Fourier transforms and TCN layers, achieving superior accuracy and interpretability in multivariate time series forecasting across diverse datasets.

Contribution

It presents the Neural Fourier Transform, a new approach that enhances interpretability and accuracy in multivariate time series forecasting, validated on multiple datasets.

Findings

Outperforms existing models on 14 datasets

Sets new benchmarks in forecasting accuracy

Offers a more interpretable forecasting model

Abstract

Multivariate time series forecasting is a pivotal task in several domains, including financial planning, medical diagnostics, and climate science. This paper presents the Neural Fourier Transform (NFT) algorithm, which combines multi-dimensional Fourier transforms with Temporal Convolutional Network layers to improve both the accuracy and interpretability of forecasts. The Neural Fourier Transform is empirically validated on fourteen diverse datasets, showing superior performance across multiple forecasting horizons and lookbacks, setting new benchmarks in the field. This work advances multivariate time series forecasting by providing a model that is both interpretable and highly predictive, making it a valuable tool for both practitioners and researchers. The code for this study is publicly available.