根据数据特性自动选最优模型，提升异构时间序列预测精度
FAME: Forecastability-Aware Mixture of Experts for Heterogeneous Time Series Forecasting
- 用多维特征指纹表征数据，智能匹配最适合的预测专家
- 在5000+设备上实现12.4%误差降低，平均每条数据仅调用1.92个专家
- 适合大规模零售/工业场景，可替代人工选型，部署已落地
大规模零售与工业预测系统包含大量异构时间序列，其生命周期、稀疏性、波动性、季节性、频谱模式和上下文敏感性差异显著。单一模型难以在所有场景下表现优异，而密集集成会增加推理开销且无法揭示专家适用性。本文研究可预测性感知的专家路由：学习数据特征如何决定预测专家的适用性。提出方法 method{}，一种稀疏混合专家框架，通过多维可预测性指纹表示每条序列，从验证性能中挖掘专家适用性目标，并训练一个成本敏感的稀疏路由器，为每条序列激活少量预算内的专家。基于山东新北洋（SNBC）的生产级售货机销售数据集（5000+机器，6000万+交易），结合公开零售基准，结果显示专家适用性在不同数据域中系统性变化。在工业数据集上， method{} Top-2 相比最强单个专家 LightGBM，均方误差降低 12.4%，平均每个序列仅激活 1.92 个专家。部署组件生成需求预测，库存收益通过固定补货策略的离线回放模拟估算。该框架将异构销售预测从启发式模型选择转变为可预测性模式的数据挖掘与专家专业化。
原文摘要 · Abstract (English)
Large-scale retail and industrial forecasting systems contain many heterogeneous time series whose lifecycle, sparsity, volatility, seasonality, spectral patterns, and contextual sensitivity differ substantially. A single forecasting model rarely performs well across all regimes, while dense ensembles increase inference cost and provide limited insight into expert suitability. This paper studies forecastability-aware expert routing: learning how data characteristics determine the suitability of forecasting experts. We propose \method{}, a sparse mixture-of-experts framework that represents each series with a multidimensional forecastability fingerprint, mines expert-suitability targets from validation performance, and trains a cost-aware sparse router to activate a small budgeted set of experts for each series. Using a production-scale vending-machine sales dataset from Shandong New Beiyang (SNBC), where the forecasting component has been integrated into the replenishment-planning pipeline, together with public retail benchmarks, we show that expert suitability varies systematically across data regimes. On the industrial dataset with 5,000+ machines and 60M+ transactions, \method{} Top-2 reduces MSE by 12.4\% over the strongest single expert, LightGBM, while executing 1.92 experts per series on average. The deployed component produces demand forecasts, while inventory-oriented gains are estimated by an offline replay simulator under a fixed replenishment policy rather than by online intervention. The framework turns heterogeneous sales forecasting from heuristic model selection into data mining of forecastability patterns and expert specialization. Code is available at https://github.com/hit636/FAME
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。