Summary of Timeseriesbench: An Industrial-grade Benchmark For Time Series Anomaly Detection Models, by Haotian Si et al.
TimeSeriesBench: An Industrial-Grade Benchmark for Time Series Anomaly Detection Models
by Haotian Si, Jianhui Li, Changhua Pei, Hang Cui, Jingwen Yang, Yongqian Sun, Shenglin Zhang, Jingjing Li, Haiming Zhang, Jing Han, Dan Pei, Gaogang Xie
First submitted to arxiv on: 16 Feb 2024
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 The proposed paper addresses the limitations of current time series anomaly detection (TSAD) methods by introducing a unified model that can be used to detect anomalies in large-scale systems with tens of thousands of curves, and evaluating its performance on newly incoming unseen time series. The authors also propose an industrial-grade benchmark called TimeSeriesBench, which assesses the performance of existing algorithms across over 168 evaluation settings and provides comprehensive analysis for future design of anomaly detection algorithms. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper proposes a new approach to detect anomalies in time series data by using a single model that can be applied to any number of curves. This is different from current methods, which require training a separate model for each curve. The authors also evaluate the performance of existing algorithms on newly incoming unseen time series and provide a benchmark dataset to test future algorithms. |
Keywords
* Artificial intelligence * Anomaly detection * Time series
