Summary of Timexer: Empowering Transformers For Time Series Forecasting with Exogenous Variables, by Yuxuan Wang et al.
TimeXer: Empowering Transformers for Time Series Forecasting with Exogenous Variables
by Yuxuan Wang, Haixu Wu, Jiaxiang Dong, Guo Qin, Haoran Zhang, Yong Liu, Yunzhong Qiu, Jianmin Wang, Mingsheng Long
First submitted to arxiv on: 29 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 proposed TimeXer model enhances time series forecasting by incorporating exogenous variables, which provide valuable external information for endogenous variables. Unlike traditional multivariate and univariate approaches that treat all variables equally or ignore exogenous information, TimeXer reconciles endogenous and exogenous data through patch-wise self-attention and variate-wise cross-attention. This approach learns global endogenous tokens to bridge causal relationships between exogenous and endogenous series. Experimental results demonstrate state-of-the-art performance on twelve real-world benchmarks, showcasing the model’s generality and scalability. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary TimeXer is a new way to improve time series forecasting by using information from external variables. This helps make predictions more accurate because it takes into account things that happen outside of what you’re trying to predict. The approach uses special techniques to combine this external information with the information from what you’re trying to predict, making it better at guessing what will happen next. |
Keywords
* Artificial intelligence * Cross attention * Self attention * Time series
