aiqbench analysis
FAME introduces a cost-aware sparse routing mechanism that dynamically selects a small subset of experts based on a series' forecastability fingerprint, significantly reducing inference overhead while improving accuracy on heterogeneous industrial datasets. By shifting from heuristic model selection to data-driven expert specialization, this approach effectively balances performance gains with computational efficiency in large-scale forecasting pipelines. The method demonstrates that activating only the top two experts can outperform the strongest single model, offering a practical solution for managing diverse time series in production environments.
FAME: Forecastability-Aware Mixture of Experts for Heterogeneous Time Series Forecasting
arXiv:2606.08896v1 Announce Type: new Abstract: 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
Abstract: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 this https URL
Artificial Intelligence (cs.AI)
arXiv:2606.08896 [cs.AI]
(or arXiv:2606.08896v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2606.08896 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 314 words
Related coverage
Beyond Static Positional Encoding: Bayesian Time-aware Attention for Non-Stationary Long-term Time Series Forecasting
arXiv · Oct 15, 2026
Personalized Federated Sparse Adaptation of Time-Series Foundation Models
arXiv · Aug 6, 2026
Structure Aware Neural Architecture Search for Mixture of Experts
arXiv · Sep 1, 2026
Fuzzy-MoE: Interpretable Regime-Conditioned Expert Routing for Non-Stationary Multivariate Time Series Forecasting
arXiv · Aug 24, 2026