Summary of Expand and Compress: Exploring Tuning Principles For Continual Spatio-temporal Graph Forecasting, by Wei Chen et al.
Expand and Compress: Exploring Tuning Principles for Continual Spatio-Temporal Graph Forecasting
by Wei Chen, Yuxuan Liang
First submitted to arxiv on: 16 Oct 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 prompt tuning-based continuous forecasting method is proposed to address the dual challenges of spatio-temporal forecasting in streaming scenarios. The method follows two fundamental tuning principles: expand and compress, which resolve inefficiency in retraining models and catastrophic forgetting over long-term history. The approach integrates a base spatio-temporal graph neural network with a continuous prompt pool, utilizing stored prompts to optimize the model for sequential learning from data streams. Experimental results on multiple real-world datasets demonstrate the superiority of this method over state-of-the-art baselines. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A new way to predict traffic flow and other things that happen in space and time is being developed. Right now, predicting these things is hard because we don’t have a good way to use all the data from sensors that are constantly sending us information. The new method gets better over time by learning from all the data it receives, not just what’s new. This makes it really useful for tasks like predicting traffic flow and air quality. |
Keywords
» Artificial intelligence » Graph neural network » Prompt
