Ensemble Classifiers and Feature-Based Methods for Structural Damage Assessment

Table 4

Classification accuracy of the selected feature subset obtained by using different feature extraction methods for bridge health monitoring benchmark problem (case C).

Signal processing technique	Optimal feature set	Accuracy (%)	Optimal feature set	Accuracy (%)
	Sensor N2		Sensor N1
STFT	S5	95.2	S5, S10, S3, S6	75.6
Fast-ST	S5, S2	63.7	S5, S7	60.5
WT	S3, S4	60.1	S7, S4, S8, S9	57.3
SWT	S5	76.4	S5, S4	58.5
EWT	S8, S5, S1, S2	71.2	S3, S8, S4	82.4