An Efficient and Effective Model to Handle Missing Data in Classification

Table 2

Accuracies achieved by simulation under MCAR missing mechanism.

Variable	Model	Complete Variable	Missing Percent									Exclude Variable
			10%	20%	30%	40%	50%	60%	70%	80%	90%
	BART.m	66.56	66.27	65.97	65.66	65.34	65.05	64.80	64.50	64.22	63.97	63.65
	BART.i	66.56	66.04	65.44	64.75	63.99	63.29	62.56	61.80	61.24	60.86	63.65
	RF.i	63.63	63.32	62.94	62.36	61.80	61.22	60.60	60.03	59.49	59.15	60.74
	BART.m	66.56	66.55	66.56	66.50	66.52	66.54	66.56	66.57	66.57	66.65	66.67
	BART.i	66.56	66.53	66.42	66.32	66.20	65.89	65.61	65.11	64.38	62.69	66.67
	RF.i	63.65	63.61	63.47	63.30	63.14	62.82	62.60	62.23	61.62	60.26	65.91
	BART.m	66.56	66.14	65.71	65.30	64.89	64.56	64.21	63.77	63.44	63.11	62.62
	BART.i	66.56	65.91	64.97	63.90	62.85	61.64	60.64	59.48	58.51	57.16	62.62
	RF.i	63.63	63.23	62.56	61.79	61.01	60.16	59.30	58.34	57.47	56.16	63.03
	BART.m	66.56	65.91	65.21	64.53	63.88	63.30	62.65	62.01	61.42	60.83	60.14
	BART.i	66.56	65.67	64.49	63.21	61.86	60.52	59.28	57.91	56.61	55.15	60.14
	RF.i	63.63	63.08	62.26	61.28	60.28	59.24	58.17	56.97	55.85	54.47	59.60