Robust Anomaly Recognition in Hydraulic Structural Safety Monitoring: A Methodology Based on Deconfounding Boosted Regression Trees

Table 5

Anomaly detection accuracy of models in different scenarios.

Model	Scenario 1	Scenario 2	Scenario 3	Average
Causal model	0.862	0.862	0.862	0.862
Noncausal model	0.936	0.824	0.745	0.835
TSBRT	0.917	0.912	0.908	0.912
CDBRT	0.922	0.919	0.904	0.915