TAT: Temporal-Aligned Transformer for Multi-Horizon Peak Demand Forecasting

2025-07-14

AI Summary

Accurate multi-step-ahead peak demand forecasting remains challenging in high-risk scenarios—such as e-commerce flash sales and brick-and-mortar retail promotions—due to the complex, event-driven nature of demand surges. Method: This paper proposes a Temporal Alignment Transformer (TAT) that explicitly incorporates prior contextual information (e.g., promotions, holidays) via a novel Temporal Alignment Attention mechanism within an encoder-decoder architecture. This mechanism enables dynamic, context-aware alignment between exogenous events and historical time series, thereby capturing event-triggered peak patterns more precisely. The model also explicitly integrates known external variables to enhance temporal structure modeling. Contribution/Results: Evaluated on a large-scale real-world e-commerce dataset, TAT achieves up to a 30% improvement in peak demand prediction accuracy and consistently outperforms state-of-the-art time series forecasting models across standard metrics.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Planning, Routing, and Scheduling: Temporal Planning
• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Multi-horizon time series forecasting has many practical applications such as demand forecasting. Accurate demand prediction is critical to help make buying and inventory decisions for supply chain management of e-commerce and physical retailers, and such predictions are typically required for future horizons extending tens of weeks. This is especially challenging during high-stake sales events when demand peaks are particularly difficult to predict accurately. However, these events are important not only for managing supply chain operations but also for ensuring a seamless shopping experience for customers. To address this challenge, we propose Temporal-Aligned Transformer (TAT), a multi-horizon forecaster leveraging apriori-known context variables such as holiday and promotion events information for improving predictive performance. Our model consists of an encoder and decoder, both embedded with a novel Temporal Alignment Attention (TAA), designed to learn context-dependent alignment for peak demand forecasting. We conduct extensive empirical analysis on two large-scale proprietary datasets from a large e-commerce retailer. We demonstrate that TAT brings up to 30% accuracy improvement on peak demand forecasting while maintaining competitive overall performance compared to other state-of-the-art methods.

Problem

• Accurate multi-horizon peak demand forecasting for supply chains
• Predicting demand during high-stake sales events effectively
• Leveraging context variables like holidays for improved forecasts

Innovation

• Temporal-Aligned Transformer for demand forecasting
• Leverages holiday and promotion event data
• Uses Temporal Alignment Attention mechanism