LLM4TS: Aligning Pre-Trained LLMs as Data-Efficient Time-Series Forecasters

系列(地层学) 时间序列 计算机科学 机器学习 地质学 古生物学
作者
Ching Chang,Wen-Chih Peng,Tien-Fu Chen
出处
期刊:Cornell University - arXiv 被引量:8
标识
DOI:10.48550/arxiv.2308.08469
摘要

Multivariate time-series forecasting is vital in various domains, e.g., economic planning and weather prediction. Deep train-from-scratch models have exhibited effective performance yet require large amounts of data, which limits real-world applicability. Recently, researchers have leveraged the representation learning transferability of pre-trained Large Language Models (LLMs) to handle limited non-linguistic datasets effectively. However, incorporating LLMs with time-series data presents challenges of limited adaptation due to different compositions between time-series and linguistic data, and the inability to process multi-scale temporal information. To tackle these challenges, we propose LLM4TS, a framework for time-series forecasting with pre-trained LLMs. LLM4TS consists of a two-stage fine-tuning strategy: the \textit{time-series alignment} stage to align LLMs with the nuances of time-series data, and the \textit{forecasting fine-tuning} stage for downstream time-series forecasting tasks. Furthermore, our framework features a novel two-level aggregation method that integrates multi-scale temporal data within pre-trained LLMs, enhancing their ability to interpret time-specific information. In experiments across 7 time-series forecasting datasets, LLM4TS is superior to existing state-of-the-art methods compared with trained-from-scratch models in full-shot scenarios, and also achieves an average improvement of 6.84% in MSE in few-shot scenarios. In addition, evaluations compared with different self-supervised learning approaches highlight LLM4TS's effectiveness with representation learning in forecasting tasks.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
sunhhhh完成签到 ,获得积分10
1秒前
4秒前
抽象的脆脆完成签到,获得积分10
4秒前
二甲双胍完成签到,获得积分20
4秒前
liull发布了新的文献求助10
6秒前
等待青枫发布了新的文献求助10
10秒前
斯文败类应助lishihao采纳,获得10
10秒前
11秒前
苹果白凡完成签到,获得积分10
12秒前
时尚初柳完成签到,获得积分10
13秒前
yiyi发布了新的文献求助20
14秒前
陈槊诸发布了新的文献求助10
16秒前
16秒前
善良的碧灵完成签到 ,获得积分10
17秒前
科研通AI5应助44采纳,获得10
18秒前
ding应助留胡子的海采纳,获得10
20秒前
20秒前
20秒前
许甜甜鸭完成签到,获得积分10
21秒前
bkagyin应助洋子采纳,获得10
23秒前
23秒前
lishihao发布了新的文献求助10
23秒前
称心凡完成签到,获得积分10
23秒前
24秒前
yyb1993完成签到 ,获得积分10
24秒前
XxxxxxENT完成签到,获得积分10
25秒前
阿童木完成签到,获得积分10
25秒前
寡妇哥完成签到 ,获得积分10
25秒前
攀攀完成签到,获得积分10
25秒前
波比发布了新的文献求助10
26秒前
天道酬勤完成签到,获得积分20
27秒前
曦cherish发布了新的文献求助10
27秒前
zyyyyyyyyyyy完成签到,获得积分10
28秒前
28秒前
JC发布了新的文献求助10
29秒前
30秒前
blueisthe完成签到 ,获得积分10
30秒前
Wxxxxx完成签到 ,获得积分10
30秒前
黎乐荷发布了新的文献求助10
31秒前
33秒前
高分求助中
Encyclopedia of Mathematical Physics 2nd edition 888
Technologies supporting mass customization of apparel: A pilot project 600
Hydropower Nation: Dams, Energy, and Political Changes in Twentieth-Century China 500
Introduction to Strong Mixing Conditions Volumes 1-3 500
Pharmacological profile of sulodexide 400
Optical and electric properties of monocrystalline synthetic diamond irradiated by neutrons 320
共融服務學習指南 300
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 物理 生物化学 纳米技术 计算机科学 化学工程 内科学 复合材料 物理化学 电极 遗传学 量子力学 基因 冶金 催化作用
热门帖子
关注 科研通微信公众号,转发送积分 3805315
求助须知:如何正确求助?哪些是违规求助? 3350274
关于积分的说明 10348210
捐赠科研通 3066165
什么是DOI,文献DOI怎么找? 1683589
邀请新用户注册赠送积分活动 809064
科研通“疑难数据库(出版商)”最低求助积分说明 765214