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

Table 4

Classification accuracy (%) of different and combinations for three datasets.

	0.1/0.9	0.2/0.8	0.3/0.7	0.4/0.6	0.5/0.5	0.6/0.4	0.7/0.3	0.8/0.2	0.9/0.1
SMB
k = 20	84.96	85.33	90.50	88.56	89.92	85.96	73.97	75.10	70.53
k = 40	79.92	87.84	89.47	89.08	86.52	87.07	84.85	79.47	67.38
k = 80	85.44	88.21	89.92	89.63	89.64	86.82	80.73	76.52	68.09
GEM
k = 20	75.87	82.11	86.06	90.81	87.35	90.07	89.92	79.96	79.46
k = 40	86.04	87.79	86.82	88.35	87.92	86.32	83.90	75.91	70.55
k = 80	83.81	84.91	83.59	87.77	87.48	84.72	86.57	79.73	79.92
STAR
k = 20	77.26	67.29	78.64	78.18	80.78	83.69	85.85	87.56	86.77
k = 40	69.61	69.09	77.27	78.76	82.06	82.21	82.14	85.88	84.97
k = 80	69.72	65.36	75.36	71.55	82.03	81.81	85.75	87.91	87.23