# Foundation Models for Demand Forecasting via Dual-Strategy Ensembling

arXiv: https://arxiv.org/abs/2507.22053 · HTML: https://mlsanity.com/paper/2507.22053/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 14 of its 52 references that are in the map.
- First posted on arXiv 29 Jul 2025.

## Builds on (most cited first, 14 of 14)

- [Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting](https://mlsanity.com/paper/2012.07436/) (2020)
- [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://mlsanity.com/paper/2211.14730/) (2022)
- [DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks](https://mlsanity.com/paper/1704.04110/) (2017)
- [One Fits All:Power General Time Series Analysis by Pretrained LM](https://mlsanity.com/paper/2302.11939/) (2023)
- [Unified Training of Universal Time Series Forecasting Transformers](https://mlsanity.com/paper/2402.02592/) (2024)
- [Efficient Processing of Deep Neural Networks: A Tutorial and Survey](https://mlsanity.com/paper/1703.09039/) (2017)
- [Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Experts](https://mlsanity.com/paper/2409.16040/) (2024)
- [TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignment](https://mlsanity.com/paper/2406.01638/) (2024)
- [TimeDiT: General-purpose Diffusion Transformers for Time Series Foundation Model](https://mlsanity.com/paper/2409.02322/) (2024)
- [TS-Reasoner: Domain-Oriented Time Series Inference Agents for Reasoning and Automated Analysis](https://mlsanity.com/paper/2410.04047/) (2024)
- [M5 Competition Uncertainty: Overdispersion, distributional forecasting, GAMLSS and beyond](https://mlsanity.com/paper/2107.06675/) (2021)
- [A Dynamic Approach to Stock Price Prediction: Comparing RNN and Mixture of Experts Models Across Different Volatility Profiles](https://mlsanity.com/paper/2410.07234/) (2024)
- [MCDFN: Supply Chain Demand Forecasting via an Explainable Multi-Channel Data Fusion Network Model](https://mlsanity.com/paper/2405.15598/) (2024)
- [Exploring the representativeness of the M5 competition data](https://mlsanity.com/paper/2103.02941/) (2021)

## Built on by (most cited first, 5 of 5)

- [A Survey of Reasoning and Agentic Systems in Time Series with Large Language Models](https://mlsanity.com/paper/2509.11575/) (2025)
- [TemporalBench: A Benchmark for Evaluating LLM-Based Agents on Contextual and Event-Informed Time Series Tasks](https://mlsanity.com/paper/2602.13272/) (2026)
- [A Unified Framework for Knowledge Transfer in Bidirectional Model Scaling](https://mlsanity.com/paper/2603.07506/) (2026)
- [Speaking Numbers to LLMs: Multi-Wavelet Number Embeddings for Time Series Forecasting](https://mlsanity.com/paper/2606.26487/) (2026)
- [Verification of Adaptive Agentic Controllers through Finite Rule Revision](https://mlsanity.com/paper/2607.09770/) (2026)

## Abstract (arXiv)

Accurate demand forecasting is critical for supply chain optimization, yet remains difficult in practice due to hierarchical complexity, domain shifts, and evolving external factors. While recent foundation models offer strong potential for time series forecasting, they often suffer from architectural rigidity and limited robustness under distributional change. In this paper, we propose a unified ensemble framework that enhances the performance of foundation models for sales forecasting in real-world supply chains. Our method combines two complementary strategies: (1) Hierarchical Ensemble (HE), which partitions training and inference by semantic levels (e.g., store, category, department) to capture localized patterns; and (2) Architectural Ensemble (AE), which integrates predictions from diverse model backbones to mitigate bias and improve stability. We conduct extensive experiments on the M5 benchmark and three external sales datasets, covering both in-domain and zero-shot forecasting. Results show that our approach consistently outperforms strong baselines, improves accuracy across hierarchical levels, and provides a simple yet effective mechanism for boosting generalization in complex forecasting environments.