Summary of Rpmixer: Shaking Up Time Series Forecasting with Random Projections For Large Spatial-temporal Data, by Chin-chia Michael Yeh et al.
RPMixer: Shaking Up Time Series Forecasting with Random Projections for Large Spatial-Temporal Data
by Chin-Chia Michael Yeh, Yujie Fan, Xin Dai, Uday Singh Saini, Vivian Lai, Prince Osei Aboagye, Junpeng Wang, Huiyuan Chen, Yan Zheng, Zhongfang Zhuang, Liang Wang, Wei Zhang
First submitted to arxiv on: 16 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 A novel approach to spatial-temporal forecasting is introduced in this paper, which leverages general time series forecasting models rather than explicitly incorporating spatial relationships. The authors propose an all-Multi-Layer Perceptron (all-MLP) architecture called RPMixer, which builds upon the success of all-MLP models in time series forecasting benchmarks. By integrating random projection layers and identity mapping residual connections, the model enhances diversity among its blocks’ outputs, ultimately improving performance. Experimental results on large-scale spatial-temporal forecasting datasets demonstrate that RPMixer outperforms alternative methods, including both graph-based and general forecasting approaches. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper explores a new way to forecast things in space and time using simple computer models. These models don’t directly use the relationships between different places or times. Instead, they try to predict what will happen by looking at patterns within each place and time separately. The authors created a special model called RPMixer that combines these patterns to make better predictions. They tested this model on lots of real-world data and found it was better than other methods at forecasting things like weather, traffic, or population growth. |
Keywords
* Artificial intelligence * Time series
