TABLE 3. Classification Performance of RF-SMA-SVM Based on Eight Feature Sets.

No.	Feature subset	ACC	MCC	Sensitivity	Specificity
No.1	{Age}	0.783	0.577	0.650	0.867
No.2	{Age, LY%}	0.807	0.640	0.683	0.900
No.3	{Age, LY%, LY}	0.787	0.573	0.617	0.900
No.4	{Age, LY%, LY, N/L}	0.810	0.641	0.733	0.867
No.5	{Age, LY%, LY, N/L, NEU%}	0.843	0.704	0.800	0.867
No.6	{Age, LY%, LY, N/L, NEU%, P/L}	0.900	0.812	0.850	0.942
No.7	{Age, LY%, LY, N/L, NEU%, P/L, RDW}	0.887	0.801	0.833	0.933
No.8	{Age, LY%, LY, N/L, NEU%, P/L, RDW, EOS%}	0.843	0.688	0.750	0.900