![[2602.13939] An Adaptive Model Selection Framework for Demand Forecasting under Horizon-Induced Degradation to Support Business Strategy and Operations](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.13939] An Adaptive Model Selection Framework for Demand Forecasting under Horizon-Induced Degradation to Support Business Strategy and Operations
Summary
This article presents an adaptive model selection framework for demand forecasting, addressing challenges posed by horizon-induced degradation in business environments.
Why It Matters
In dynamic business settings with variable demand, accurate forecasting is crucial for operational efficiency. This framework enhances model selection processes, enabling businesses to make informed decisions based on specific demand conditions and forecasting horizons, ultimately improving strategic outcomes.
Key Takeaways
- The AHSIV framework improves model selection for demand forecasting.
- It addresses horizon-induced ranking instability in forecasting models.
- Empirical evaluations show AHSIV's effectiveness across multiple datasets.
- The framework integrates various error metrics and structural demand classifications.
- Model selection should be adaptive rather than static for better accuracy.
Computer Science > Machine Learning arXiv:2602.13939 (cs) [Submitted on 15 Feb 2026] Title:An Adaptive Model Selection Framework for Demand Forecasting under Horizon-Induced Degradation to Support Business Strategy and Operations Authors:Adolfo González, Víctor Parada View a PDF of the paper titled An Adaptive Model Selection Framework for Demand Forecasting under Horizon-Induced Degradation to Support Business Strategy and Operations, by Adolfo Gonz\'alez and 1 other authors View PDF Abstract:Business environments characterized by structural demand intermittency, high variability, and multi-step planning horizons require robust and reproducible model selection mechanisms. Empirical evidence shows that no forecasting model is universally dominant and that relative rankings vary across error metrics, demand regimes, and forecast horizons, generating ambiguity in multi-SKU decision contexts. This study proposes AHSIV (Adaptive Hybrid Selector for Intermittency and Variability), a horizon-aware and regime-conditioned model selection framework designed to address horizon-induced ranking instability. The proposed approach integrates scaled and absolute error metrics adjusted through a Metric Degradation by Forecast Horizon (MDFH) procedure, structural demand classification, multi-objective Pareto dominance, and hierarchical bias refinement within a unified decision architecture. The empirical evaluation is conducted on the Walmart, M3, M4, and M5 datasets under multiple train-t...
