Good morning,
can anyone help me? How we can use page rank algorithm for detection of anomalies ?
Maybe this books will give an answer on your problems: Application of Modified PageRank Algorithm for Anomaly Detection in Movements of Older Adults
A good way to look for anomalies is to run page rank and then examine the nodes that are "outliers" -- meaning anything more than 3 standard deviations away from the mean, or even just the highest percentiles. If you'd like to get a sense of the distribution of your data, you can run pagerank in the .stats
mode with CALL gds.pageRank.stats
-- see the docs for more details - PageRank - Neo4j Graph Data Science
2 Likes