A Path-Based Feature Selection Algorithm for Enterprise Credit Risk Evaluation

Table 3

Classification accuracy (%) comparison for three datasets.

	Ours	FS-RRC	mIMR	mRMR	RCDFS	FCBF
SMB
k = 20	90.50	90.52	90.47	89.47	90.35	89.92
k = 40	89.47	86.95	87.34	88.64	89.18	88.63
k = 80	89.92	89.54	87.39	88.78	89.20	89.11
GEM
k = 20	90.81	89.05	90.46	88.81	91.25	90.06
k = 40	88.35	86.79	84.45	86.52	85.83	86.70
k = 80	87.77	87.71	86.72	85.81	86.45	84.74
STAR
k = 20	87.56	87.47	90.10	85.50	87.03	87.52
k = 40	85.88	85.83	82.92	83.08	84.47	85.17
k = 80	87.91	87.35	86.06	84.86	87.71	84.73