Summary of Cross-city Few-shot Traffic Forecasting Via Traffic Pattern Bank, by Zhanyu Liu et al.
Cross-city Few-Shot Traffic Forecasting via Traffic Pattern Bank
by Zhanyu Liu, Guanjie Zheng, Yanwei Yu
First submitted to arxiv on: 17 Aug 2023
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 proposes a novel framework, called Traffic Pattern Bank (TPB), to improve traffic forecasting in cities with limited data availability. The authors recognize that current approaches rely heavily on data-rich cities, which can’t be generalized to data-scarce cities. To address this challenge, TPB leverages pre-trained encoders and clustering techniques to generate a bank of traffic patterns across different cities. These patterns are then used to guide spatial-temporal models in forecasting future traffic. The framework also incorporates meta-training using Reptile to fine-tune initial parameters. Experimental results on real-world datasets demonstrate the effectiveness of TPB in cross-city few-shot traffic forecasting, outperforming existing methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us predict traffic better by sharing knowledge between cities with lots of data and those with not so much. Right now, we rely too heavily on data-rich cities, which isn’t fair to the ones that don’t have as many sensors or devices. The authors came up with a clever idea called Traffic Pattern Bank (TPB) that uses special computer algorithms to group similar traffic patterns from different cities together. This allows us to use the knowledge gained in one city to predict traffic in another, even if it has fewer data points. By combining this pattern bank with a model that looks at both space and time, we can make better predictions about what’s going to happen on the roads. |
Keywords
* Artificial intelligence * Clustering * Few shot
