Summary of Attractor Memory For Long-term Time Series Forecasting: a Chaos Perspective, by Jiaxi Hu et al.
Attractor Memory for Long-Term Time Series Forecasting: A Chaos Perspective
by Jiaxi Hu, Yuehong Hu, Wei Chen, Ming Jin, Shirui Pan, Qingsong Wen, Yuxuan Liang
First submitted to arxiv on: 18 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Chaotic Dynamics (nlin.CD)
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 This paper presents a novel model for long-term time series forecasting called Attraos, which incorporates chaos theory to capture the complex dynamics of real-world data. The authors recognize that traditional models often ignore the underlying continuous dynamic systems that generate discrete time series and instead attempt to model their chaotic nature. Attraos uses non-parametric Phase Space Reconstruction embedding and a multi-scale dynamic memory unit to learn historical dynamics and predict future values using a frequency-enhanced local evolution strategy. Theoretical analysis and empirical evidence demonstrate that Attraos outperforms existing methods on mainstream LTSF datasets and chaotic datasets, with fewer parameters than PatchTST. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about creating a new way to forecast long-term time series data. Right now, many models try to predict what will happen next without considering the underlying systems that create the data in the first place. The authors of this paper think that these systems are chaotic and unpredictable, like weather patterns or stock market trends. They created a model called Attraos that tries to capture these dynamics by learning from historical data. It’s like trying to understand how a river flows over time. The results show that Attraos is better than other models at predicting the future, and it uses fewer calculations to do so. |
Keywords
* Artificial intelligence * Embedding * Time series
