Table 3.

Scenario 2 (decreasing global trend mean): Comparison of hot-spot detection under small and large hot-spots with four criterions: precision, recall, F-measure and ${\mathrm{ARL}}_1$.

	Large shift $\delta /\sigma =5$				Small shift $\delta /\sigma =1$
Methods	Precision	Recall	F-measure	ARL	Precision	Recall	F-measure	${\mathrm{ARL}}_1$
SSD-Tensor	0.3242	0.9978	0.6610	1.3490	0.3245	0.9978	0.6612	3.4120
	(0.0184)	(0.0085)	(0.0105)	(0.4762)	(0.0188 )	(0.0085)	(0.0108)	(0.8217)
Scan-stat	0.3111	0.4666	0.3889	–	0.2567	0.3851	0.3209	–
	(0.0007)	(0.0011)	(0.0009)	(–)	(0.0307)	(0.0460)	(0.0383)	(–)
ZQ-LASSO	0.0000	0.0000	0.0000	50.0000	0.0000	0.0000	0.0000	50.0000
	(0.0000)	(0.0000)	(0.0000)	(0.0000)	(0.0000)	(0.0000)	(0.0000)	(0.0000)
PCA	–	–	–	50.0000	–	–	–	50.0000
	–	–	–	(0.0000)	–	–	–	(0.0000)
T2	–	–	–	50.0000	–	–	–	50.0000
	–	–	–	(0.0000)	–	–	–	(0.0000)