/graph_anomaly_detect

Anomaly detection in time evolving graphs using the NetSimile algorithm

Primary LanguagePython

##Anomaly detection in time evolving graphs Finding anomalies in graphs using NetSimile[1] algorithm

Local Configuration

After cloning the repo on your local machine, add a file named 'dev_settings.py' in your repo which has the following contents:

DATA_URL=""
SAVE_URL=""

The path mentioned in 'DATA_URL' should contain all the unzipped data in 4 folders:

as-733  enron  p2p-Gnutella  reality_mining_voices

Generated Plots

####AS-733 Plot

####Enron Plot ####p2p-Gnutella Plot ####Reality Mining Voices Plot

References

[1] Berlingerio, Michele, et al. "NetSimile: a scalable approach to size-independent network similarity." arXiv preprint arXiv:1209.2684 (2012).