[论文解读] Zero-shot Forecasting by Simulation Alone

Boris N. Oreshkin, Mayank Jauhari+11 more|arXiv (Cornell University)|Jan 2, 2026

Forecasting Techniques and Applications被引用 0

一句话总结

本论文提出 SarSim0，一种基于 SARIMA 的快速时间序列模拟器，通过在纯合成数据上对神经预测器进行预训练实现零-shot 预测，在 M-Series 和 GiftEval 基准上展现出较强的泛化能力。结果显示，合成数据在预训练方面可以与真实数据相媲美，甚至在某些零-shot 情况下优于特定基线。

ABSTRACT

Zero-shot time-series forecasting holds great promise, but is still in its infancy, hindered by limited and biased data corpora, leakage-prone evaluation, and privacy and licensing constraints. Motivated by these challenges, we propose the first practical univariate time series simulation pipeline which is simultaneously fast enough for on-the-fly data generation and enables notable zero-shot forecasting performance on M-Series and GiftEval benchmarks that capture trend/seasonality/intermittency patterns, typical of industrial forecasting applications across a variety of domains. Our simulator, which we call SarSim0 (SARIMA Simulator for Zero-Shot Forecasting), is based off of a seasonal autoregressive integrated moving average (SARIMA) model as its core data source. Due to instability in the autoregressive component, naive SARIMA simulation often leads to unusable paths. Instead, we follow a three-step procedure: (1) we sample well-behaved trajectories from its characteristic polynomial stability region; (2) we introduce a superposition scheme that combines multiple paths into rich multi-seasonality traces; and (3) we add rate-based heavy-tailed noise models to capture burstiness and intermittency alongside seasonalities and trends. SarSim0 is orders of magnitude faster than kernel-based generators, and it enables training on circa 1B unique purely simulated series, generated on the fly; after which well-established neural network backbones exhibit strong zero-shot generalization, surpassing strong statistical forecasters and recent foundation baselines, while operating under strict zero-shot protocol. Notably, on GiftEval we observe a "student-beats-teacher" effect: models trained on our simulations exceed the forecasting accuracy of the AutoARIMA generating processes.

研究动机与目标

• 在现实数据稀缺、偏倚或易泄露的工业场景中，推动零-shot 预测的研究。
• 开发一个快速、无泄露风险的合成数据生成器，以大规模预训练预测模型。
• 以稳定的 SARIMA 动力学为基础，并扩展以支持多季节性与重尾噪声。
• 证明在无目标数据微调的情况下，使用合成数据预训练的模型对异质基准具有良好泛化能力。

提出的方法

• 以 SARIMA 作为合成序列的数据生成核心。
• 通过将极点限制在单位圆内并从极点表征推导系数来稳定化仿真。
• 引入 SARIMA-2，通过基础过程与包络过程的叠加或乘法耦合来捕捉双季节性。
• 附加 Noiser 模块以注入重尾、水平相关扰动（泊松、广义伽马和对数正态分布）。
• 实现跨多条轨迹的向量化生成，以实现对数十亿序列的即时合成。
• 仅在 SarSim0 生成的数据上训练基础模型骨架（如 NBEATS、PatchTST、Chronos-Small T5），并在基准上评估零-shot 性能。

Figure 1: SarSim0 simulator pipeline. Top: two base components are generated by SARIMA with AR (and seasonal) roots sampled via the characteristic polynomial inside the stability region, yielding well-behaved paths at seasonalities $s\!=\!24$ and $s\!=\!7$ . Middle: a SARIMA-2 superposition/modulati

实验结果

研究问题

• RQ1 基于 SARIMA 的合成数据生成器是否能产生可用于训练预测模型的现实时间序列模式？
• RQ2 仅在合成数据上进行预训练是否能在多样的真实基准上实现强大的零-shot 泛化？
• RQ3 在同一合成数据上预训练时，不同架构的归纳偏置（如 NBEATS、PatchTST、Chronos）的表现如何？
• RQ4 各个仿真组件（SARIMA、SARIMA-2、Noisers）对零-shot 预测性能的贡献有多大？

主要发现

• 在异质基准上，经过 SarSim0 训练的模型实现了强大的零-shot 泛化，甚至超越了一些真实数据预训练的基线。
• 在 SarSim0 上预训练的模型常常接近具有大量真实数据预训练的模型，甚至在某些合成基线（如 KernelSynth 和 ForecastPFN）上表现优于。
• 拥有多样归纳偏置（密集、基于注意力、补丁化）的架构在同一合成数据上训练后也能取得具有竞争力的性能，显示对模型选择的鲁棒性。
• 在 GiftEval 上，基于 SarSim0 预训练的模型展现出“学生胜过教师”的效应，优于 AutoARIMA 生成的过程。
• 消融研究表明 SARIMA-2 和 Noisers 对泛化能力有实质性贡献，尤其是 SARIMA-2 对不同骨干模型的准确性尤为重要。

Figure 2: Sampling of SARIMA poles by SarSim0 . The SARIMA order-10 AR process poles are shown along with the unit circle on the left. The resulting generated processes with these poles are shown on the right. The top pane shows poles sampled according to the proposed procedure, resulting in a reali