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

Table 1 Comparison of Naive Bayes classifier and alternative approach: marginal probability of class variable = 0.3

Model	Number of predictors	Sample size	Sensitivity	Specificity	Positive predictive value	Negative predictive value	Precision
NB	5	500	39.21	86.81	58.13	76.05	58.13
AODE	5	500	36.78	88.82	60.64	75.71	60.64
TAN	5	500	38.17	87.81	59.31	75.93	59.31
NB-BLCA (EM)	5	500	35.46	90.69	64.13	75.69	64.13
NB	5	1000	38.98	86.7	57.76	75.96	57.76
AODE	5	1000	35.8	88.76	59.8	75.43	59.8
TAN	5	1000	37.61	87.61	58.56	75.73	58.56
NB-BLCA (EM)	5	1000	33.25	90.74	62.45	75.06	62.45
NB	5	2000	38.75	86.64	57.56	75.86	57.56
AODE	5	2000	35.71	88.51	59.23	75.32	59.23
TAN	5	2000	37.36	87.49	58.24	75.61	58.24
NB-BLCA (EM)	5	2000	31.74	90.81	61.35	74.6	61.35
NB	10	500	49.48	87.13	63.48	79.37	63.48
AODE	10	500	49.86	88.27	65.81	79.7	65.81
TAN	10	500	49.92	87.69	64.73	79.61	64.73
NB-BLCA (EM)	10	500	66.38	92.52	80.08	86.04	80.08
NB	10	1000	48.47	87.26	63.19	79.06	63.19
AODE	10	1000	48.34	88.1	64.71	79.18	64.71
TAN	10	1000	48.51	87.51	63.66	79.12	63.66
NB-BLCA (EM)	10	1000	59.43	91.62	76.19	83.46	76.19
NB	10	2000	48.49	87.1	62.94	79	62.94
AODE	10	2000	48.12	87.74	63.97	79.01	63.97
TAN	10	2000	48.41	87.2	63.08	79	63.08
NB-BLCA (EM)	10	2000	54.27	90.92	72.96	81.58	72.96
NB	20	500	59.13	89.33	71.18	83.12	71.18
AODE	20	500	62.15	90.31	74.12	84.32	74.12
TAN	20	500	60.33	89.8	72.52	83.61	72.52
NB-BLCA (EM)	20	500	98.73	99.69	99.31	99.43	99.31
NB	20	1000	57.66	89.26	70.54	82.57	70.54
AODE	20	1000	59.26	89.82	72.21	83.2	72.21
TAN	20	1000	57.91	89.4	70.91	82.67	70.91
NB-BLCA (EM)	20	1000	97.63	99.33	98.48	98.95	98.48
NB	20	2000	57.29	89.12	70.04	82.47	70.04
AODE	20	2000	58.14	89.44	70.96	82.81	70.96
TAN	20	2000	57.31	89.13	70.07	82.48	70.07
NB-BLCA (EM)	20	2000	95.86	98.8	97.24	98.18	97.24

ISSN: 1471-2288