Developing Machine Learning and Statistical Tools to Evaluate the Accessibility of Public Health Advice on Infectious Diseases among Vulnerable People

Table 5

Thresholds, positive/negative likelihood ratio, and 95% CI of the best-performing GNB classifier on the test data.

Probability thresholds	Sensitivity (95% CI)	Specificity (95% CI)	Positive likelihood ratio (LR+) (95% CI)	Negative likelihood ratio (LR-) (95% CI)
0.1	0.982 (0.95, 1.0)	0.9118 (0.82, 1.0)	11.12 (3.77, 32.79)	0.02031 (0.003, 0.142)
0.2	0.982 (0.95, 1.0)	0.941 (0.86, 1.0)	16.69 (4.35, 64.04)	0.01968 (0.003, 0.137)
0.5	0.963 (0.91, 1.0)	0.941 (0.86, 1.0)	16.37 (4.26, 62.87)	0.039 (0.010, 0.154)
0.6	0.963 (0.91, 1.0)	1.0 (1.0, 1.0)	Infinity	0.037 (0.010, 0.144)
0.7	0.9444 (0.883, 1.0)	1.0 (1.0, 1.0)	Infinity	0.0556 (0.019, 0.167)
0.9	0.9074 (0.830, 0.985)	1.0 (1.0, 1.0)	Infinity	0.09259 (0.0402, 0.213)
Machine learning classifier: refined GNB classifier (with ALFCW and Nh features)