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

TL;DR

This paper introduces TAT, a novel transformer-based model that leverages context variables like holidays and promotions to improve multi-horizon peak demand forecasting accuracy, especially during high-stake sales events.

Contribution

The paper proposes the Temporal-Aligned Transformer (TAT) with a new Temporal Alignment Attention mechanism for better context-dependent demand prediction.

Findings

Up to 30% accuracy improvement on peak demand forecasting.

Effective handling of high-stake sales events.

Competitive overall performance compared to state-of-the-art methods.

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.