发表机构
Amirkabir University of Technology（伊朗德黑兰阿米尔卡比尔理工大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究针对机器学习需求预测运营波动性大的问题，提出上下文反卷积（CD）方法，通过核调制算子分离冲击与基线、分层部分池化实现目录规模部署，经样本外评估，CD在降低成本和提高预测可靠性上有贡献。
AI 中文摘要
机器学习需求预测优化了统计准确性，但留下了过多的运营波动性，这会增加安全库存并放大牛鞭效应。我们引入了上下文反卷积（CD），这是一种两阶段估计器，它将需求感知重新构建为凸分解：一个核调制带状算子将瞬态促销驱动的冲击与平滑的结构基线分开，并且分层部分池化使得无需每个SKU训练即可进行目录规模的部署。该算子是数据驱动的，而不是强加的——在促销响应是脉冲式的地方（M5的大部分，Favorita的全部）它简化为恒等算子，并且仅在存在真正的多日结转的地方起作用，因此收益基于结构分解本身。在30,490个M5 SKU和2,845个Favorita商品上进行严格的样本外评估，给定CD相同的未来日历并具有日历感知基线，我们基于完整的库存成本核算来确定贡献：CD降低了安全库存、持有成本和订单方差，但对事件峰值供应不足，仅在持有成本超过缺货成本的约20%（95%置信区间[17%,25%]）时才降低总成本；否则它是一个运营稳定性和库存资本层，而不是预期成本最小化器。其准确性贡献在于可靠性而非集中趋势：在十一个基线中，CD实现了每个SKU误差的最低横截面离散度，并且在0.8%的SKU上错误预测比每个基线减少了200%以上，而每个基线为9.9 - 20.6%，在所有四个M5抽取中均排名第一。由于任何足够平滑的预测都会使方差比和基于标准差的安全库存最小化，我们将它们视为诊断指标，而不是目标。一项支持性分析表明，学习到的需求算子是非正态的，但CD紧凑的参数核以可解释的方式匹配了它们的运营性能。
英文摘要
Machine learning demand forecasts optimize statistical accuracy yet leave excess operational volatility that inflates safety stock and amplifies the Bullwhip effect. We introduce \textbf{Contextual Deconvolution} (CD), a two-stage estimator that reframes demand sensing as a convex decomposition: a kernel-modulated banded operator separates transient promotion-driven shocks from a smooth structural baseline, and hierarchical partial pooling enables catalog-scale deployment without per-SKU training. The operator is data-derived, not imposed---it reduces to the identity wherever the promotional response is impulsive (most of M5, all of Favorita) and contributes only where genuine multi-day carryover exists, so the gains rest on the structural decomposition itself. Evaluating strictly out-of-sample on 30,490 M5 SKUs and 2,845 Favorita items, with calendar-aware baselines given CD's identical future calendar, we anchor the contribution on a full inventory-cost accounting: CD lowers safety stock, holding cost, and order variance but under-provisions event spikes, reducing total cost only when holding costs exceed $\sim$20\% of stockout costs (95\% CI $[17\%,25\%]$); otherwise it is an operational-stability and inventory-capital layer, not an expected-cost minimizer. Its accuracy contribution is reliability rather than central tendency: across eleven baselines, CD attains the lowest cross-sectional dispersion of per-SKU error and mis-forecasts by more than 200\% on 0.8\% of SKUs versus 9.9--20.6\% for every baseline, ranking first on both in all four M5 draws. Because the Variance Ratio and std-based safety stock are minimized by any sufficiently smooth forecast, we treat them as diagnostics, not objectives. A supporting analysis shows the learned demand operators are non-normal, yet CD's compact parametric kernel matches their operational performance interpretably.
Comments46 pages, 12 figures, 22 tables
