Summary of Matrix Profile For Anomaly Detection on Multidimensional Time Series, by Chin-chia Michael Yeh et al.
Matrix Profile for Anomaly Detection on Multidimensional Time Series
by Chin-Chia Michael Yeh, Audrey Der, Uday Singh Saini, Vivian Lai, Yan Zheng, Junpeng Wang, Xin Dai, Zhongfang Zhuang, Yujie Fan, Huiyuan Chen, Prince Osei Aboagye, Liang Wang, Wei Zhang, Eamonn Keogh
First submitted to arxiv on: 14 Sep 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Databases (cs.DB)
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 explores anomaly detection in multidimensional time series data, a common challenge in real-world applications such as manufacturing factories. The Matrix Profile (MP), previously effective in univariate time series anomaly detection, becomes complex when dealing with multidimensional scenarios. To address this, the authors investigate strategies for condensing the pairwise distance tensor into a profile vector and extend MP to efficiently find k-nearest neighbors for anomaly detection. Benchmarking against 19 baseline methods on 119 datasets across three learning setups (unsupervised, supervised, and semi-supervised), the paper demonstrates that MP consistently delivers high performance. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how to detect unusual patterns in data collected from multiple sensors at a factory or other place where things are happening over time. Right now, there’s no good way to do this with lots of different types of data coming in. The authors want to make the “Matrix Profile” method work better for this kind of data. They try different ways to make it work and then test it against other methods on a bunch of real-world datasets. What they find is that their method does really well across all kinds of learning situations. |
Keywords
» Artificial intelligence » Anomaly detection » Semi supervised » Supervised » Time series » Unsupervised
