ReNF: Rethinking the Design of Neural Long-Term Time Series Forecasters

Neural Forecasters (NFs) have become a cornerstone of Long-term Time Series Forecasting (LTSF). However, recent progress has been hampered by an overemphasis on architectural complexity at the expense of fundamental forecasting structures. In this work, we revisit principled designs of LTSF. We begin by formulating a Variance Reduction Hypothesis (VRH), positing that generating and combining multiple forecasts is essential to reducing the inherent uncertainty of NFs. Guided by this, we propose Boosted Direct Output (BDO), a streamlined paradigm that synergistically hybridizes the causal structure of Auto-Regressive (AR) with the stability of Direct Output (DO), while implicitly realizing the principle of forecast combination within a single network. Furthermore, we mitigate a critical validation-test generalization gap by employing parameter smoothing to stabilize optimization. Extensive experiments demonstrate that these trivial yet principled improvements enable a direct temporal MLP to outperform recent, complex state-of-the-art models in nearly all benchmarks, without relying on intricate inductive biases. Finally, we empirically verify our hypothesis, establishing a dynamic performance bound that highlights promising directions for future research. The code is publicly available at: \url{https://github.com/Luoauoa/ReNF}.

This research highlights the critical importance of foundational design over architectural complexity in long-term time series forecasting (LTSF). We demonstrate that combining stabilization techniques (such as weight averaging) with an improved forecasting structure, in the spirit of classic forecast combination, enables a simple MLP to outperform heavily parameterized state-of-the-art models. This work encourages the machine learning community to reconsider fundamental forecasting principles rather than relying solely on architectural engineering. Furthermore, demonstrating the efficacy of simple, well-regularized models promotes highly computationally efficient and environmentally friendly deployment of industrial forecasting systems.