Suggest for faster calculation of scores

caisr-hh / group-anomaly-detection

GRAND: Group-based Anomaly Detection for Large-Scale Monitoring of Complex Systems

MIT License

15 stars 9 forks source link

Suggest for faster calculation of scores #5

Closed agiannoul closed 8 months ago

agiannoul commented 4 years ago

Suggest for faster calculation of scores in Reference Group when knn mean distance is selected for non-conformity metric using KDTree from sklearn.neighbors library.