A Bayesian latent class extension of naive Bayesian classifier and its application to the classification of gastric cancer patients

Table 2 Comparison of Naive Bayes classifier and alternative approach: marginal probability of class variable = 0.5

Model	Number of predictors	Sample size	Sensitivity	Specificity	Positive predictive value	Negative predictive value	Precision
NB	5	500	64.5	67.25	66.69	65.72	66.69
AODE	5	500	65.28	67.51	67.16	66.33	67.16
TAN	5	500	64.96	67.43	66.97	66.08	66.97
NB-BLCA (EM)	5	500	65.75	68.97	68.32	67.07	68.32
NB	5	1000	64.36	67.06	66.39	65.71	66.39
AODE	5	1000	64.73	67.24	66.69	66.02	66.69
TAN	5	1000	64.63	67.07	66.51	65.87	66.51
NB-BLCA (EM)	5	1000	64.93	68.12	67.25	66.33	67.25
NB	5	2000	64.39	66.82	66.34	65.59	66.34
AODE	5	2000	64.43	67.1	66.57	65.73	66.57
TAN	5	2000	64.34	66.96	66.42	65.61	66.42
NB-BLCA (EM)	5	2000	64.39	67.69	66.77	65.75	66.77
NB	10	500	70.26	71.57	71.31	70.67	71.31
AODE	10	500	71.08	72.65	72.33	71.57	72.33
TAN	10	500	70.81	72.26	71.97	71.24	71.97
NB-BLCA (EM)	10	500	79.43	82.61	82.16	80.19	82.16
NB	10	1000	70.08	71.25	71	70.45	71
AODE	10	1000	70.58	71.84	71.57	70.98	71.57
TAN	10	1000	70.28	71.41	71.17	70.64	71.17
NB-BLCA (EM)	10	1000	76.36	79.27	78.75	77.12	78.75
NB	10	2000	69.95	70.95	70.67	70.34	70.67
AODE	10	2000	70.26	71.26	71	70.66	71
TAN	10	2000	69.98	71.01	70.73	70.38	70.73
NB-BLCA (EM)	10	2000	73.8	76.62	75.96	74.64	75.96
NB	20	500	75.97	76.79	76.62	76.22	76.62
AODE	20	500	77.43	78.88	78.6	77.82	78.6
TAN	20	500	76.77	77.59	77.43	77.02	77.43
NB-BLCA (EM)	20	500	99.01	99.36	99.36	99.01	99.36
NB	20	1000	75.43	76.19	76.03	75.63	76.03
AODE	20	1000	76.24	77.34	77.11	76.51	77.11
TAN	20	1000	75.7	76.42	76.27	75.89	76.27
NB-BLCA (EM)	20	1000	98.13	98.62	98.61	98.15	98.61
NB	20	2000	75.39	75.99	75.89	75.52	75.89
AODE	20	2000	75.83	76.64	76.49	76	76.49
TAN	20	2000	75.41	76.02	75.92	75.54	75.92
NB-BLCA (EM)	20	2000	96.92	97.58	97.56	96.94	97.56

ISSN: 1471-2288