Early warning systems for malaria outbreaks in Thailand: an anomaly detection approach

Malaria Journal

Table 1 Methods used for anomaly detection and their references

Method	References
Statistical profiling	[12, 38,39,40,41,42]
Predictive confidence interval	[42,43,44,45]
Unsupervised clustering¹	[12, 16, 37, 46, 47]
Weekly case comparison	[4] 06/01/2024 06:55:00
Monthly case comparison	[48]
Rolling historical average²
Weekly 3 year median	[11]

¹Two main unsupervised clustering techniques are used for this method: time-series clustering with tsclust [37] and density-based clustering using DBSCAN
²This method was developed based on a combination of techniques from statistical profiling and weekly case comparison

ISSN: 1475-2875