Imprecise Imputation as a Tool for Solving Classification Problems with Mean Values of Unobserved Features

Table 4

The ACC measures for real datasets by different values of 𝑛.

Dataset	𝑛	ACC1	ACC2	ACC3	ACC4
Iris	20	0.961	0.982	0.977	0.998
	40	0.944	0.963	0.986	0.998
Pima	20	0.501	0.598	0.502	0.573
Indian Diabetes	40	0.532	0.655	0.590	0.621
Mammographic Masses	20	0.642	0.769	0.750	0.722
	40	0.654	0.781	0.778	0.739
Parkinsons	20	0.567	0.721	0.706	0.637
	40	0.56	0.721	0.713	0.644
Breast Cancer Wisconsin (Original)	20	0.885	0.923	0.937	0.963
	40	0.832	0.919	0.815	0.962
Breast Cancer Wisconsin (Diagnostic)	20	0.820	0.835	0.885	0.921
	40	0.785	0.851	0.825	0.926
Musk	20	0.565	0.648	0.625	0.629
	40	0.606	0.678	0.689	0.658
Indian Liver	20	0.508	0.674	0.666	0.614
Patients	40	0.519	0.682	0.660	0.621
Lung-Cancer	12	0.647	0.709	0.693	0.637
	16	0.666	0.679	0.637	0.786