AI 中文总结
针对冻结时间序列预测器，提出FreshCast框架，通过持续更新非参数记忆并校准权重，在七个基准和十种架构上平均降低MSE 14.6%和5.6%，验证了及时且能纠正残差的历史参考的重要性。
AI 中文摘要
检索增强的时间序列预测使用与当前上下文相似的历史片段的延续作为预测器的参考。大多数现有方法从训练片段中一次性构建检索记忆，使得部署后观测到的数据无法作为参考，并且通常不校准检索到的信息应如何影响冻结的预测器。我们确定了检索对冻结预测器效用的两个关键决定因素：历史是否仍然反映当前状态，以及其引发的修正是否与预测器的残差误差对齐，这种对齐在记忆变得过时后可能在验证和部署之间发生偏移。我们提出FreshCast，一个即插即用的检索框架，保持预测器冻结，用新观测值持续更新非参数记忆，通过关系核回归形成记忆预测，并在验证片段上以封闭形式校准其权重。在一个简化的生成模型下，我们通过预测器误差与记忆修正之间的二阶关系刻画了最优组合增益，并表明足够长的回看窗口可以使周期性记忆信息变得冗余。在七个基准和十种预测架构上，FreshCast在输入长度为96和720时，对每个评估的预测器和输入长度平均降低MSE 14.6%和5.6%，并在其比较设置中实现了比评估的检索增强和在线基线更低的MSE。消融实验表明，在训练结束时冻结记忆会消除大部分增益，从而确定训练后观测是改进的主要来源。对于冻结的预测器，有用的历史参考必须保持及时，并提供有助于纠正其剩余误差的信息。
英文摘要
Retrieval-augmented time-series forecasting uses the continuations of historical segments similar to the current context as references for a forecaster. Most existing methods build the retrieval memory once from the training segment, leaving observations revealed after deployment unavailable as references, and generally do not calibrate how much the retrieved information should influence a frozen forecaster. We identify two key determinants of retrieval utility for a frozen forecaster: whether the history still reflects the current state, and whether the correction it induces aligns with the forecaster's residual errors, an alignment that can shift between validation and deployment when the memory becomes stale. We propose FreshCast, a plug-in retrieval framework that keeps the forecaster frozen, continuously updates a non-parametric memory with new observations, forms a memory forecast through relational kernel regression, and calibrates its weight in closed form on the validation segment. Under a simplified generative model, we characterize the optimal combination gain through the second-order relation between forecaster error and memory correction, and show that a sufficiently long look-back can make periodic memory information redundant. Across seven benchmarks and ten forecasting architectures, FreshCast reduces average MSE for every evaluated forecaster and input length, by 14.6% and 5.6% at input lengths 96 and 720, and achieves lower MSE than the evaluated retrieval-augmented and online baselines in their comparison settings. Ablations show that freezing the memory at the end of training removes most of the gain, identifying post-training observations as a primary source of improvement. For a frozen forecaster, useful historical references must remain timely and provide information that helps correct its remaining errors.
Comments13 pages, 6 figures, 6 tables, Work in progress
