Comparing machine learning algorithms for predicting COVID-19 mortality

Table 4 Features affecting predicting mortality in patients with COVID-19

Row	Features name	Degree of importance	Row	Features name	Degree of importance
1	Dyspnea	0.5532	21	Chest pain and pressure	0.2256
2	ICU admission	0.5409	22	Absolute neutrophil count	0.2123
3	Oxygen therapy	0.3789	23	Headache	0.1992
4	Age	0.3207	24	Gender	0.186
5	Fever	0.3142	25	Gastrointestinal symptoms	0.1802
6	Cough	0.3072	26	White cell count	0.1702
7	Loss of taste	0.2944	27	C-reactive protein	0.1574
8	Loss of smell	0.2923	28	Hypersensitive troponin	0.1428
9	Hypertension	0.2768	29	Pneumonia	0.1066
10	Contusion	0.2744	30	Glucose	0.0906
11	Muscular Pain	0.2731	31	Erythrocyte sedimentation rate	0.0826
12	Chill	0.2537	32	Creatinine	0.0716
13	Runny noise	0.2532	33	Alkaline phosphatase	0.0678
14	Blood urea nitrogen	0.2524	34	Length of hospitalization	0.0626
15	Diabetes	0.2506	35	Aspartate aminotransferase	0.0445
16	Sore throat	0.25	36	Smoking	0.0427
17	Absolute lymphocyte count	0.2339	37	Alanine aminotransferase	0.0319
18	Nausea/vomiting	0.2301	38	Platelet count	0.0210
19	Other under line disease	0.2282	39
20	Cardiac disease	0.2274	39

ISSN: 1472-6947