Summary of Autotimes: Autoregressive Time Series Forecasters Via Large Language Models, by Yong Liu et al.
AutoTimes: Autoregressive Time Series Forecasters via Large Language Models
by Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long
First submitted to arxiv on: 4 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computation and Language (cs.CL)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper proposes a novel approach to time series forecasting, leveraging large language models (LLMs) to repurpose them as autoregressive time series forecasters. The model, called AutoTimes, projects time series into the embedding space of language tokens and autoregressively generates future predictions with arbitrary lengths. This allows for flexibility in lookback length and scalability with larger LLMs. Additionally, the paper introduces textual timestamps to align multivariate time series. Empirical results show that AutoTimes achieves state-of-the-art performance with only 0.1% trainable parameters and over 5training/inference speedup compared to advanced LLM-based forecasters. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper develops a new way to predict future events in time series data, like stock prices or weather patterns. It uses special computer models called large language models (LLMs) that were originally designed for understanding human language. The new approach, called AutoTimes, is very good at predicting the future and does it much faster than other methods. It also allows experts to easily adjust how far back in time they look when making predictions. This could be very useful for many different types of forecasting. |
Keywords
* Artificial intelligence * Autoregressive * Embedding space * Inference * Time series
