TimeGPT-1

TL;DR

TimeGPT is a pioneering foundation model for time series that achieves accurate, efficient, and simple predictions across diverse datasets, leveraging AI insights to advance the field.

Contribution

It introduces the first large-scale foundation model for time series, demonstrating zero-shot inference capabilities and cross-domain applicability.

Findings

TimeGPT outperforms traditional methods in accuracy and efficiency.

Zero-shot inference with TimeGPT is effective across various datasets.

Large-scale models can democratize access to precise time series predictions.

Abstract

In this paper, we introduce TimeGPT, the first foundation model for time series, capable of generating accurate predictions for diverse datasets not seen during training. We evaluate our pre-trained model against established statistical, machine learning, and deep learning methods, demonstrating that TimeGPT zero-shot inference excels in performance, efficiency, and simplicity. Our study provides compelling evidence that insights from other domains of artificial intelligence can be effectively applied to time series analysis. We conclude that large-scale time series models offer an exciting opportunity to democratize access to precise predictions and reduce uncertainty by leveraging the capabilities of contemporary advancements in deep learning.