TimeHF: Billion-Scale Time Series Models Guided by Human Feedback

TL;DR

TimeHF introduces a billion-parameter time series model guided by human feedback, significantly improving prediction accuracy in industrial supply chain applications.

Contribution

The paper presents a novel pipeline for large time series models with human feedback, including new embedding and optimization techniques.

Findings

Improved prediction accuracy by 33.21% in supply chain tasks.

Successfully deployed in JD.com's supply chain for automated replenishment.

Demonstrates scalability and industrial relevance of large time series models.

Abstract

Time series neural networks perform exceptionally well in real-world applications but encounter challenges such as limited scalability, poor generalization, and suboptimal zero-shot performance. Inspired by large language models, there is interest in developing large time series models (LTM) to address these issues. However, current methods struggle with training complexity, adapting human feedback, and achieving high predictive accuracy. We introduce TimeHF, a novel pipeline for creating LTMs with 6 billion parameters, incorporating human feedback. We use patch convolutional embedding to capture long time series information and design a human feedback mechanism called time-series policy optimization. Deployed in JD.com's supply chain, TimeHF handles automated replenishment for over 20,000 products, improving prediction accuracy by 33.21% over existing methods. This work advances LTM technology and shows significant industrial benefits.