analiza aiqbench
This study advances intermittent-demand forecasting by unifying the trade-off between retaining item-specific history and borrowing strength from shared priors within a hierarchical empirical-Bayes framework. The authors introduce a diagnostic that quantifies when pooling yields tangible accuracy gains, revealing that excessive truncation can degrade performance by raising the noise floor against which cross-series patterns must be resolved. By demonstrating significant cost reductions while maintaining near-optimal accuracy on sparse datasets, this work provides a rigorous theoretical and practical guide for scaling forecasting models to massive, low-frequency series.
When Does Pooling Pay? Credibility and Resolution under Forgetting in Intermittent-Demand Forecasting
arXiv:2511.12749v3 Announce Type: replace-cross Abstract: Forecasting many sparse series requires two choices: how much of each series' own past to retain, and how much to borrow from other series. We show that when one exponential recency operator is applied to item- and group-level statistics alike in a hierarchical empirical-Bayes hurdle model, the two become one decision: forgetting preserves the leverage of a shared prior while raising the noise floor against which fine cross-series structure must be resolved. We formalize this two-sided effect as a fitting-window diagnostic, a credibility bound and a resolution condition for candidate pools, and derive from it an ex ante screen for regimes where refinement cannot pay. On five public intermittent-demand panels comprising over 19,000 series, where item-level evidence is sparsest, the diagnostic is right in both directions. Where truncation lowers credibility the shared prior pays: switching it off costs $5\%$ and $12\%$ at the shortest histories on the two panels it helps most, against about $1\%$ at full length, and the gain sits in the occurrence block. On four panels the diagnostic reports at most $0.6\%$ of room for refinement and realized gains are no larger; on the one panel with room the learned mixture does not collect it, which locates the open problem in the partition objective. The resulting forecaster, EBB, stays within about $3\%$ of a per-series Gaussian-process specialist at fixed origin and within $1\%$ under walk-forward evaluation where both run, at two to three orders of magnitude lower cost.
Abstract:Forecasting many sparse series requires two choices: how much of each series' own past to retain, and how much to borrow from other series. We show that when one exponential recency operator is applied to item- and group-level statistics alike in a hierarchical empirical-Bayes hurdle model, the two become one decision: forgetting preserves the leverage of a shared prior while raising the noise floor against which fine cross-series structure must be resolved. We formalize this two-sided effect as a fitting-window diagnostic, a credibility bound and a resolution condition for candidate pools, and derive from it an ex ante screen for regimes where refinement cannot pay. On five public intermittent-demand panels comprising over 19,000 series, where item-level evidence is sparsest, the diagnostic is right in both directions. Where truncation lowers credibility the shared prior pays: switching it off costs $5\%$ and $12\%$ at the shortest histories on the two panels it helps most, against about $1\%$ at full length, and the gain sits in the occurrence block. On four panels the diagnostic reports at most $0.6\%$ of room for refinement and realized gains are no larger; on the one panel with room the learned mixture does not collect it, which locates the open problem in the partition objective. The resulting forecaster, EBB, stays within about $3\%$ of a per-series Gaussian-process specialist at fixed origin and within $1\%$ under walk-forward evaluation where both run, at two to three orders of magnitude lower cost.
Preprint. 52 pages, 15 figures. Equal contribution by the two authors. v3: Retitled and substantially revised; reframes the work around forgetting and pooling, with credibility/resolution theory and expanded experiments
Machine Learning (stat.ML); Machine Learning (cs.LG)
arXiv:2511.12749 [stat.ML]
(or arXiv:2511.12749v3 [stat.ML] for this version)
Text integral sub licența deschisă a publicației · 284 cuvinte
Subiecte conexe
[AINews] Meta Muse Wonder & Wonder Full: MSL’s Claude 5.5-tier Frontier Model
Latent Space · 13 oct. 2026
MARGIN: Runtime Confidence Calibration for Multi-Agent Foundation Model Coordination
arXiv · 3 aug. 2026
Benchmarking and Training Reward Models for Cross-Source Evidence Grounding in Search Agents
arXiv · 15 oct. 2026
How to apply to AI safety fellowships (and beyond)
LessWrong · 2 oct. 2026