Summary of Graph Spatiotemporal Process For Multivariate Time Series Anomaly Detection with Missing Values, by Yu Zheng et al.
Graph Spatiotemporal Process for Multivariate Time Series Anomaly Detection with Missing Values
by Yu Zheng, Huan Yee Koh, Ming Jin, Lianhua Chi, Haishuai Wang, Khoa T. Phan, Yi-Ping Phoebe Chen, Shirui Pan, Wei Xiang
First submitted to arxiv on: 11 Jan 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 paper introduces GST-Pro, a novel framework for detecting anomalies in multivariate time series data. This framework tackles the challenges posed by missing values and irregularly-sampled observations. The approach consists of two main components: a graph spatiotemporal process based on neural controlled differential equations, which models spatial and temporal dependencies even with missing values; and an anomaly scoring mechanism that alleviates reliance on complete uniform observations. Experimental results show GST-Pro outperforms state-of-the-art methods in detecting anomalies. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Anomalies in time series data are important to detect for practical applications like power grids and traffic forecasting. Current approaches struggle because real-world data is often unstructured, with missing values and irregular sampling. The paper presents GST-Pro, a new way to detect anomalies that can handle these challenges. It uses a special process called graph spatiotemporal process and a scoring system to find unusual patterns. The results show that GST-Pro works well and is better than other methods. |
Keywords
* Artificial intelligence * Spatiotemporal * Time series
