An Efficient and Effective Model to Handle Missing Data in Classification

Table 4

Accuracies achieved by simulation under MNAR missing mechanism.

Variable	Model	Complete Variable	Missing Percent									Exclude Variable
			10%	20%	30%	40%	50%	60%	70%	80%	90%
	BART.m	66.56	66.35	66.16	65.87	65.51	65.10	64.76	64.44	64.16	63.92	63.65
	BART.i	66.56	66.04	65.43	64.70	64.15	63.70	63.47	63.39	63.33	63.11	63.65
	RF.i	63.63	63.27	62.86	62.31	61.75	61.40	60.99	60.73	60.57	60.27	60.74
	BART.m	66.56	66.57	66.52	66.55	66.55	66.54	66.55	66.59	66.58	66.65	66.67
	BART.i	66.56	66.51	66.46	66.36	66.20	65.99	65.57	65.10	64.27	62.93	66.67
	RF.i	63.65	63.62	63.47	63.35	63.14	62.94	62.58	62.19	61.60	60.51	65.91
	BART.m	66.56	66.07	65.52	65.04	64.61	64.20	63.85	63.53	63.31	63.05	62.62
	BART.i	66.56	65.51	64.38	63.30	62.25	61.30	60.38	59.47	58.70	57.45	62.62
	RF.i	63.63	62.91	62.11	61.36	60.58	59.83	59.09	58.35	57.62	56.53	63.03
	BART.m	66.56	66.07	65.44	64.87	64.21	63.50	62.90	62.20	61.55	60.92	60.14
	BART.i	66.56	65.32	63.89	62.61	61.24	59.94	58.82	57.58	56.49	55.28	60.14
	RF.i	63.63	62.73	61.66	60.72	59.64	58.59	57.63	56.69	55.63	54.47	59.60