FAME: Forecastability-Aware Mixture of Experts for Heterogeneous Time Series Forecasting
Researchers have proposed FAME, a forecastability-aware sparse mixture-of-experts framework designed to route heterogeneous time series to the most suitable forecasting models based on data characteristics. The system was tested on a production-scale vending machine dataset from Shandong New Beiyang (SNBC) comprising over 5,000 machines and 60 million transactions, as well as public retail benchmarks. FAME reduced mean squared error by 12.4% over the strongest single expert model while activating fewer than two experts per series on average, suggesting a path toward more efficient and interpretable large-scale forecasting.
FAME (Forecastability-Aware Mixture of Experts) addresses a core challenge in industrial forecasting: a single model rarely performs well across time series that differ in sparsity, volatility, seasonality, and other characteristics, while dense ensembles are computationally expensive. The framework represents each time series with a multidimensional 'forecastability fingerprint,' mines expert-suitability targets from validation performance, and trains a cost-aware sparse router to activate only a small budgeted set of experts per series. In experiments on SNBC's vending machine sales data, the FAME Top-2 configuration reduced MSE by 12.4% compared to LightGBM, the strongest individual expert, while averaging just 1.92 expert activations per series. The framework has been integrated into SNBC's replenishment-planning pipeline, though inventory-level gains were estimated via an offline replay simulator under a fixed replenishment policy rather than through live online intervention. The authors argue the approach transforms heterogeneous sales forecasting from heuristic model selection into a systematic data-mining problem of forecastability patterns and expert specialization, with code made publicly available.
What's missing
The generalizability of the forecastability fingerprint and routing mechanism to domains outside retail and vending machines (e.g., energy, finance) is not demonstrated. The paper does not report statistical significance tests for the 12.4% MSE improvement, nor does it detail how the expert pool was selected or whether the results are sensitive to that choice.
What different sources said
- arXiv cs.AICenter
FAME: Forecastability-Aware Mixture of Experts for Heterogeneous Time Series Forecasting
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.