发表机构
AMSS Center of Forecasting Science, Chinese Academy of Sciences; Department of Computer Science, Northwestern University; Department of Statistics and Data Science, Northwestern University; Department of Statistics and Operations Research, University of North Carolina（中国科学院预测科学中心; 西北大学计算机科学系; 西北大学统计与数据科学系; 北卡罗来纳大学统计与运筹学系）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对多时间序列高维协变量预测难题，Hopformer提出两阶段框架，第一阶段用SPA提取共同趋势，第二阶段用LoRA微调Transformer建模残差依赖，理论上有依据，实验中提升了预测性能。
AI 中文摘要
预测具有高维协变量的多个时间序列存在核心挑战：统一常见时间模式，同时保留有意义的特定序列信息。我们引入了Hopformer（同质性追求Transformer），这是一个解决此挑战的两阶段框架。第一阶段，我们执行稀疏模式聚合（SPA）方案，提取包含协变量的共同低方差趋势，作为同质化层。第二阶段，LoRA微调的Transformer对残差中的剩余复杂依赖关系建模。我们的方法有理论依据，证明了SPA通过神谕不等式实现了近乎最优的偏差-方差权衡，还为第二阶段在相关时间序列数据下提供了泛化界。Hopformer创造了新的技术水平，在合成和现实世界预测基准上平均将MASE提高了6.56%。
英文摘要
Forecasting multiple time-series with high-dimensional covariates presents a core challenge: unifying common temporal patterns while retaining meaningful series-specific information. We introduce Hopformer (Homogeneity-Pursuit Transformer), a two-stage framework that addresses this challenge. In the first stage, we perform a Sparsity Pattern Aggregation (SPA) scheme extracting a common low-variance trend that incorporates the covariates. This acts as a homogenization layer. In the second stage, a LoRA-fine-tuned Transformer models the remaining complex dependencies in the residual. Our method is theoretically grounded. We prove that SPA achieves a near-optimal bias-variance trade-off via an oracle inequality. We also provide generalization bounds for the second stage under dependent time series data. Hopformer sets a new state of the art, improving MASE by an average of 6.56% across synthetic and real-world forecasting benchmarks.
