In-Context Fine-Tuning for Time-Series Foundation Models

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Das, Abhimanyu, Faw, Matthew, Sen, Rajat, Zhou, Yichen
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929570438971392
author Das, Abhimanyu
Faw, Matthew
Sen, Rajat
Zhou, Yichen
author_facet Das, Abhimanyu
Faw, Matthew
Sen, Rajat
Zhou, Yichen
contents Motivated by the recent success of time-series foundation models for zero-shot forecasting, we present a methodology for $\textit{in-context fine-tuning}$ of a time-series foundation model. In particular, we design a pretrained foundation model that can be prompted (at inference time) with multiple time-series examples, in order to forecast a target time-series into the future. Our foundation model is specifically trained to utilize examples from multiple related time-series in its context window (in addition to the history of the target time-series) to help it adapt to the specific distribution of the target domain at inference time. We show that such a foundation model that uses in-context examples at inference time can obtain much better performance on popular forecasting benchmarks compared to supervised deep learning methods, statistical models, as well as other time-series foundation models. Interestingly, our in-context fine-tuning approach even rivals the performance of a foundation model that is explicitly fine-tuned on the target domain.
format Preprint
id arxiv_https___arxiv_org_abs_2410_24087
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle In-Context Fine-Tuning for Time-Series Foundation Models
Das, Abhimanyu
Faw, Matthew
Sen, Rajat
Zhou, Yichen
Machine Learning
Artificial Intelligence
Computation and Language
Motivated by the recent success of time-series foundation models for zero-shot forecasting, we present a methodology for $\textit{in-context fine-tuning}$ of a time-series foundation model. In particular, we design a pretrained foundation model that can be prompted (at inference time) with multiple time-series examples, in order to forecast a target time-series into the future. Our foundation model is specifically trained to utilize examples from multiple related time-series in its context window (in addition to the history of the target time-series) to help it adapt to the specific distribution of the target domain at inference time. We show that such a foundation model that uses in-context examples at inference time can obtain much better performance on popular forecasting benchmarks compared to supervised deep learning methods, statistical models, as well as other time-series foundation models. Interestingly, our in-context fine-tuning approach even rivals the performance of a foundation model that is explicitly fine-tuned on the target domain.
title In-Context Fine-Tuning for Time-Series Foundation Models
topic Machine Learning
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2410.24087