Feature Screening via Mutual Information Learning Based on Nonparametric Density Estimation

Table 3

The proportions of and in Example 2.

Model	Parameter	cov (X)	Method
(a)			PC-SIS	1.00	0.96	0.69	0.65
			MV-SIS	1.00	1.00	0.87	0.87
			MI-SIS2	1.00	0.88	0.73	0.55
			MI-SIS	1.00	0.98	0.58	0.57
			PC-SIS	1.00	1.00	0.98	0.98
			MV-SIS	1.00	1.00	1.00	1.00
			MI-SIS2	1.00	1.00	0.96	0.96
			MI-SIS	1.00	1.00	1.00	1.00
(b)			PC-SIS	0.35	0.59	0.99	0.20
			MV-SIS	0.39	0.56	1.00	0.28
			MI-SIS2	0.46	0.64	0.99	0.29
			MI-SIS	0.99	1.00	0.94	0.93
			PC-SIS	0.83	0.80	0.91	0.57
			MV-SIS	0.42	0.48	0.98	0.25
			MI-SIS2	0.80	0.82	0.92	0.58
			MI-SIS	1.00	1.00	0.91	0.91