A process mining- deep learning approach to predict survival in a cohort of hospitalized COVID‐19 patients

Table 5 Shapley value analysis summary

Category	Time intervals
Category	6 Hr	12Hr	18Hr	24Hr	30Hr	36Hr	42Hr	48 Hr	54Hr	60Hr	66Hr	72Hr
Demographics	0.0144	0.0706	0.5983	1.014	0.0657	0.0622	0.0222	0.2034	0.0422	0.0274	0.0199	0.0698
Comorbidity	0.0044	0.0071	0.0264	0.2162	0.0126	0.0465	0.0076	0.1012	0.0087	0.0032	0.0039	0.0058
REP Events	0.0041	0.0064	0.0143	0.0201	0.0092	0.0061	0.0049	0.0041	0.0036	0.0022	0.0037	0.0044
Lab Measurement events	0.0035	0.0062	0.0092	0.0023	0.0083	0.0048	0.0048	0.0026	0.0036	0.0022	0.0035	0.0041
marking	0.0027	0.0040	0.0079	0.0023	0.0061	0.0048	0.0043	0.0025	0.0034	0.0022	0.0034	0.0033
Location events	0.0027	0.0033	0.0068	0.0023	0.0058	0.0044	0.0035	0.0023	0.0032	0.0019	0.0032	0.0033
Linear decay function (max)	0.0025	0.0030	0.0058	0.0022	0.0053	0.0039	0.0033	0.0022	0.0028	0.0015	0.0029	0.0032
Linear decay function (mean)	0.0024	0.0030	0.0055	0.0018	0.0052	0.0038	0.033	0.0022	0.0028	0.0013	0.0020	0.0029
VIT events	0.0023	0.0027	0.0053	0.0017	0.0046	0.0031	0.0025	0.0019	0.0020	0.0012	0.0018	0.0026
Token count	0.0022	0.0027	0.0044	0.0017	0.0042	0.0028	0.0023	0.0016	0.0018	0.0011	0.0017	0.0024
Logarithmic decay function (mean)	0.0018	0.0026	0.0042	0.0016	0.0038	0.0027	0.0023	0.0015	0.0017	0.0011	0.0017	0.0021
ICU Events	0.0018	0.0019	0.0026	0.0013	0.0018	0.0024	0.0022	0.0014	0.0013	0.0002	0.0011	0.0020

ISSN: 1472-6947