Implement SAND #26
Description
Implement SAND [1] anomaly detector, which measures the distance from a normal model. This normal model is incrementally updated to be able to deal with concept drift in streaming applications.
[1] Boniol, Paul, et al. "SAND: streaming subsequence anomaly detection." Proceedings of the VLDB Endowment 14.10 (2021): 1717-1729.