Table 4.

The average importance and SD for each feature for the MLP and SVM models and the balanced data sets.

Data sets and features		MLPa, mean (SD)		SVMb, mean (SD)
RT-PCRc balanced
	Gender	0.236 (0.013)	0.230 (0.013)
	Health professional	0.048 (0.005)	0.042 (0.005)
	Fever	0.262 (0.013)	0.257 (0.013)
	Sore throat	0.103 (0.006)	0.098 (0.006)
	Dyspnea	0.096 (0.007)	0.088 (0.007)
	Olfactory disorders	0.027 (0.004)	0.015 (0.004)
	Cough	0.084 (0.008)	0.078 (0.007)
	Coryza	0.025 (0.004)	0.013 (0.004)
	Taste disorders	0.031 (0.004)	0.020 (0.004)
	Headache	0.023 (0.003)	0.002 (0.003)
Rapid balanced
	Gender	0.120 (0.009)	0.117 (0.010)
	Health professional	0.033 (0.006)	0.029 (0.005)
	Fever	0.115 (0.010)	0.105 (0.011)
	Sore throat	0.012 (0.005)	0.023 (0.004)
	Dyspnea	0.177 (0.013)	0.177 (0.014)
	Olfactory disorders	0.157 (0.012)	0.149 (0.012)
	Cough	0.082 (0.009)	0.076 (0.008)
	Coryza	0.064 (0.006)	0.058 (0.006)
	Taste disorders	0.072 (0.007)	0.055 (0.006)
	Headache	0.036 (0.006)	0.028 (0.005)
Both balanced
	Gender	0.161 (0.007)	0.154 (0.007)
	Health professional	0.025 (0.004)	0.024 (0.004)
	Fever	0.207 (0.010)	0.193 (0.009)
	Sore throat	0.084 (0.005)	0.075 (0.005)
	Dyspnea	0.078 (0.006)	0.088 (0.006)
	Olfactory disorders	0.055 (0.006)	0.049 (0.005)
	Cough	0.062 (0.005)	0.071 (0.005)
	Coryza	0.035 (0.003)	0.046 (0.003)
	Taste disorders	0.071 (0.006)	0.068 (0.005)
	Headache	0.051 (0.005)	0.045 (0.005)

^aMLP: multilayer perceptron.

^bSVM: support vector machine.

^cRT-PCR: reverse transcription polymerase chain reaction.