FreDF: Learning to Forecast in the Frequency Domain
This addresses a specific issue in time series forecasting for researchers and practitioners, offering a novel approach to improve accuracy, though it is incremental as it builds on the Direct Forecast paradigm.
The paper tackles the problem of biased multi-step time series forecasting due to overlooked label autocorrelation, proposing FreDF which learns in the frequency domain to reduce bias and significantly outperforms state-of-the-art methods in experiments.
Time series modeling presents unique challenges due to autocorrelation in both historical data and future sequences. While current research predominantly addresses autocorrelation within historical data, the correlations among future labels are often overlooked. Specifically, modern forecasting models primarily adhere to the Direct Forecast (DF) paradigm, generating multi-step forecasts independently and disregarding label autocorrelation over time. In this work, we demonstrate that the learning objective of DF is biased in the presence of label autocorrelation. To address this issue, we propose the Frequency-enhanced Direct Forecast (FreDF), which mitigates label autocorrelation by learning to forecast in the frequency domain, thereby reducing estimation bias. Our experiments show that FreDF significantly outperforms existing state-of-the-art methods and is compatible with a variety of forecast models. Code is available at https://github.com/Master-PLC/FreDF.