Empowering Time Series Forecasting with LLM-Agents

TL;DR

This paper introduces DCATS, a data-centric agent that leverages metadata to improve data quality, leading to significant error reduction in time series forecasting without changing model architecture.

Contribution

The paper presents DCATS, a novel AutoML approach that focuses on data cleaning using metadata to enhance forecasting accuracy in time series data.

Findings

DCATS achieves an average 6% error reduction across models.

Data-centric AutoML improves forecasting without altering models.

Results highlight the importance of data quality in time series prediction.

Abstract

Large Language Model (LLM) powered agents have emerged as effective planners for Automated Machine Learning (AutoML) systems. While most existing AutoML approaches focus on automating feature engineering and model architecture search, recent studies in time series forecasting suggest that lightweight models can often achieve state-of-the-art performance. This observation led us to explore improving data quality, rather than model architecture, as a potentially fruitful direction for AutoML on time series data. We propose DCATS, a Data-Centric Agent for Time Series. DCATS leverages metadata accompanying time series to clean data while optimizing forecasting performance. We evaluated DCATS using four time series forecasting models on a large-scale traffic volume forecasting dataset. Results demonstrate that DCATS achieves an average 6% error reduction across all tested models and time horizons, highlighting the potential of data-centric approaches in AutoML for time series forecasting.