. 2018 Apr 4;11:227. doi: 10.1186/s13071-018-2812-x

Table 3.

Accuracy rate of the full dataset, the training (80%) and the holdout sample (20%) of Toxocara cati, Cystoisospora spp., Ancylostomatidae, Aelurostrongylus abstrusus and Giardia spp. models

Observed cases		Predicted cases for the full-data model			Predicted cases for the split-sample validation model
		Infection status^a		Percentage correct (%)	Selected cases			Unselected cases
		Infection status^a			Infection status^a		% correct	Infection status^a		% correct
		Uninfected	Infected		Uninfected	Infected	% correct	Uninfected	Infected	% correct
Toxocara cati infection status	Uninfected	874	0	100	705	0	100	169	0	100
	Infected	278	0	0	223	0	0	55	0	0
	Accuracy rate			75.9			76.0			75.4
Cystoisospora spp. infection status	Uninfected	963	0	100	741	0	100	222	0	100
	Infected	189	0	0	155	0	0	34	0	0
	Accuracy rate			83.6			82.7			86.7
Ancylostomatidae infection status	Uninfected	961	5	99.5	766	2	99.7	197	1	99.5
	Infected	173	13	7.0	134	9	6.3	40	3	7.0
	Accuracy rate			84.5			85.1			83.0
Aelurostrongylus abstrusus infection status	Uninfected	1112	0	100	878	0	100	234	0	100
	Infected	40	0	0	31	0	0	9	0	0
	Accuracy rate			96.5			96.6			96.3
Giardia spp. infection status	Uninfected	1126	0	100	907	0	100	219	0	100
	Infected	26	0	0	18	0	0	8	0	0
	Accuracy rate			97.7			98.1			96.5

^aInfection status refers to Toxocara cati, Cystoisospora spp., Ancylostomatidae, Aelurostrongylus abstrusus and Giardia spp. predicted infection status in accordance to the observed infection status of the aforementioned parasite species