Spatial cluster detection using dynamic programming

Table 4 Comparison of model selection in terms of spatial precision

Frequency range	Coverage	DP precision	GR precision	SR precision	(DP-GR) 95% confidence interval limits		p-value
Low	< 16%	0.09365	0.11775	0.07022	- 0.02412	- 0.02408	< 0.00001
	16-34%	0.28551	0.33207	0.27251	- 0.04659	- 0.04654	< 0.00001
	≥ 35%	0.57974	0.62428	0.60255	- 0.04462	- 0.04447	< 0.00001
Mid	< 16%	0.31987	0.34414	0.11212	- 0.02430	- 0.02425	< 0.00001
	16-34%	0.67969	0.66288	0.31934	0.01679	0.01683	< 0.00001
	≥ 35%	0.92269	0.86046	0.55602	0.06211	0.06235	< 0.00001
High	< 16%	0.50417	0.53018	0.15127	- 0.02604	- 0.02599	< 0.00001
	16-34%	0.84815	0.78839	0.34174	0.05974	0.05978	< 0.00001
	≥ 35%	0.91292	0.83541	0.54478	0.07740	0.07763	< 0.00001
Very high	< 16%	0.66221	0.51208	0.14081	0.15010	0.15017	< 0.00001
	16-34%	0.84584	0.44728	0.26127	0.39854	0.39859	< 0.00001
	≥ 35%	0.92306	0.66563	0.48019	0.25728	0.25758	< 0.00001

Comparison of the DP, GR, and SR spatial precisions obtained from model selection. The table shows the mean precision for each algorithm, the lower and upper 95% confidence limits for the difference in precision (DP-GR) and the associated t-test p-values

ISSN: 1472-6947