Summary of Extreme Value Modelling Of Feature Residuals For Anomaly Detection in Dynamic Graphs, by Sevvandi Kandanaarachchi et al.
Extreme Value Modelling of Feature Residuals for Anomaly Detection in Dynamic Graphs
by Sevvandi Kandanaarachchi, Conrad Sanderson, Rob J. Hyndman
First submitted to arxiv on: 8 Oct 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 paper proposes a novel method for detecting anomalies in temporal sequences of graphs, with applications in transport networks and cyber attacks. Existing methods suffer from limitations such as high false positive rates and difficulties handling variable-sized graphs and non-trivial temporal dynamics. The approach models temporal dependencies via time series analysis of graph features, removes dependencies using residuals, and classifies remaining extremes using Extreme Value Theory. Comparative evaluations show the proposed method achieves better accuracy than TensorSplat and Laplacian Anomaly Detection. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper is about finding unusual patterns in a sequence of graphs that can help detect accidents or cyber attacks. Current methods for doing this have some big problems, like lots of false alarms and trouble handling different-sized graphs and complex timing patterns. The new approach looks at the pattern of graph features over time to understand how they relate to each other, then removes any predictable parts and uses a special statistical technique to find the truly unusual events. Tests show that this method does a much better job than two other popular methods. |
Keywords
» Artificial intelligence » Anomaly detection » Time series
