$t_0$：一种带上下文预测的时间序列基础模型
$t_0$: A Time-Series Foundation Model for Forecasting with Context
浏览论文内容
中文总结 AI 辅助
本文提出时间序列基础模型$t_0$系列，通过多变量上下文和分位数预测实现零样本预测，在多个基准上取得领先性能。
中文摘要 AI 辅助
我们提出了 $t_0$，一个用于多变量上下文预测的开源权重基础模型系列。我们发布了其前两个成员：$\ exttt{t0-alpha}$ 和 $\ exttt{t0-beta}$，参数分别为102M和256M。两者都基于目标历史、过去协变量和已知未来协变量进行预测，无需针对特定任务进行重新训练。它们的Transformer层沿时间维度和变量维度交替进行注意力计算。它们通过分位数预测产生概率性预测。预训练结合了精选的公共数据与合成生成器族，这些生成器族被构建为包含协变量到目标的依赖关系。在GIFT-Eval上，$\ exttt{t0-alpha}$ 达到了0.4941的聚合CRPS，$\ exttt{t0-beta}$ 达到了0.4738的CRPS和0.6865的MASE，两者均排名第三，且与最佳零样本TSFM的差距在4.0%以内。在fev-bench上，它们的技能分数分别为42.2和46.7，后者再次排名第三，落后领先者2.0分。我们对 $\ exttt{t0-alpha}$ 进行了深入分析。已知未来协变量在30个任务上将其技能提高了6.3个百分点。报告还检查了其校准、在长时域上的滚动策略以及对缺失数据的鲁棒性。在维多利亚电力需求基准上，$\ exttt{t0-beta}$ 是使用近一年上下文的最准确模型之一。在独立的Macrocosm评估中，针对29个月的每小时ERCOT价格，两者都将滞后价格基线的MAE降低了38%。
英文摘要
We present $t_0$, a family of open-weights foundation models for forecasting with multivariate context. We release its first two members: $\texttt{t0-alpha}$ and $\texttt{t0-beta}$, respectively 102M and 256M parameters. Both condition their forecasts on target history, past covariates, and known-future covariates, without task-specific retraining. Their transformer layers alternate attention along time and across variates. They produce probabilistic forecasts through quantile predictions. Pretraining combines curated public data with synthetic generator families constructed to contain covariate-to-target dependencies. On GIFT-Eval, $\texttt{t0-alpha}$ reaches an aggregate CRPS of 0.4941, and $\texttt{t0-beta}$ a CRPS of 0.4738 and a MASE of 0.6865, third on both and within 4.0% of the best zero-shot TSFM. On fev-bench they score 42.2 and 46.7 in skill, the latter third again and 2.0 points behind the leader. We analyze $\texttt{t0-alpha}$ in depth. Known-future covariates raise its skill by 6.3 percentage points across 30 tasks. The report also examines its calibration, its rollout strategy on long horizons, and its robustness to missing data. On the Victoria electricity-demand benchmark, $\texttt{t0-beta}$ is among the most accurate models with a context of nearly a year. In an independent Macrocosm evaluation of hourly ERCOT prices over 29 months, both cut the MAE of the lagged-price baseline by 38%.
发表机构
- The Forecasting Company（预测公司）
- Macrocosm（宏观宇宙）
机构由 AI 辅助整理，请以论文原文为准。
