Time-LLM: Time Series Forecasting by Reprogramming Large Language Models

计算机科学 杠杆(统计) 人工智能 背景(考古学) 机器学习 时间序列 系列(地层学) 语言模型 模式 地理 社会科学 生物 社会学 古生物学 考古
作者
Ming Jin,Shiyu Wang,Lintao Ma,Zhixuan Chu,James Y. Zhang,Xiaoming Shi,Pin‐Yu Chen,Yuxuan Liang,Yuan-Fang Li,Shirui Pan,Qingsong Wen
出处
期刊:Cornell University - arXiv 被引量:16
标识
DOI:10.48550/arxiv.2310.01728
摘要

Time series forecasting holds significant importance in many real-world dynamic systems and has been extensively studied. Unlike natural language process (NLP) and computer vision (CV), where a single large model can tackle multiple tasks, models for time series forecasting are often specialized, necessitating distinct designs for different tasks and applications. While pre-trained foundation models have made impressive strides in NLP and CV, their development in time series domains has been constrained by data sparsity. Recent studies have revealed that large language models (LLMs) possess robust pattern recognition and reasoning abilities over complex sequences of tokens. However, the challenge remains in effectively aligning the modalities of time series data and natural language to leverage these capabilities. In this work, we present Time-LLM, a reprogramming framework to repurpose LLMs for general time series forecasting with the backbone language models kept intact. We begin by reprogramming the input time series with text prototypes before feeding it into the frozen LLM to align the two modalities. To augment the LLM's ability to reason with time series data, we propose Prompt-as-Prefix (PaP), which enriches the input context and directs the transformation of reprogrammed input patches. The transformed time series patches from the LLM are finally projected to obtain the forecasts. Our comprehensive evaluations demonstrate that Time-LLM is a powerful time series learner that outperforms state-of-the-art, specialized forecasting models. Moreover, Time-LLM excels in both few-shot and zero-shot learning scenarios.
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
刚刚
1秒前
海潮发布了新的文献求助10
2秒前
zengzeng完成签到,获得积分10
2秒前
wanci应助Kane采纳,获得10
3秒前
乐乐应助呐呐呐采纳,获得10
4秒前
自自自在发布了新的文献求助10
5秒前
李健的粉丝团团长应助LBB采纳,获得10
5秒前
liusui发布了新的文献求助10
6秒前
写个锤子完成签到,获得积分10
6秒前
大模型应助dfggg采纳,获得10
6秒前
科研通AI5应助小高采纳,获得10
7秒前
共享精神应助科研如喝水采纳,获得10
7秒前
8秒前
俄而完成签到 ,获得积分10
8秒前
lvvyy完成签到,获得积分10
8秒前
领导范儿应助科研通管家采纳,获得10
10秒前
无花果应助科研通管家采纳,获得10
10秒前
爆米花应助科研通管家采纳,获得10
10秒前
Ava应助练习者采纳,获得10
10秒前
11秒前
打打应助迅速冰岚采纳,获得30
12秒前
202483067完成签到 ,获得积分10
12秒前
13秒前
13秒前
小刺猬完成签到,获得积分10
13秒前
喜马拉雅川完成签到,获得积分10
13秒前
传奇3应助米里迷路采纳,获得10
13秒前
科研通AI5应助liusui采纳,获得30
13秒前
zxc完成签到,获得积分10
13秒前
传奇3应助HJJHJH采纳,获得10
13秒前
Kane发布了新的文献求助10
14秒前
15秒前
Akim应助dsdjsicj采纳,获得10
15秒前
15秒前
所所应助自由的石头采纳,获得30
15秒前
yyhgyg完成签到,获得积分10
16秒前
16秒前
独特的清发布了新的文献求助20
17秒前
Dolphin123发布了新的文献求助10
17秒前
高分求助中
Les Mantodea de Guyane Insecta, Polyneoptera 2500
One Man Talking: Selected Essays of Shao Xunmei, 1929–1939 (PDF!) 1000
Technologies supporting mass customization of apparel: A pilot project 450
A Field Guide to the Amphibians and Reptiles of Madagascar - Frank Glaw and Miguel Vences - 3rd Edition 400
China Gadabouts: New Frontiers of Humanitarian Nursing, 1941–51 400
The Healthy Socialist Life in Maoist China, 1949–1980 400
Walking a Tightrope: Memories of Wu Jieping, Personal Physician to China's Leaders 400
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 3786934
求助须知:如何正确求助?哪些是违规求助? 3332593
关于积分的说明 10256397
捐赠科研通 3047840
什么是DOI,文献DOI怎么找? 1672734
邀请新用户注册赠送积分活动 801549
科研通“疑难数据库(出版商)”最低求助积分说明 760271