BMC Bioinformatics

Table 4 Simulation results for model 3

From: Sparse sliced inverse regression for high dimensional data analysis

Model 3
(\(\rho\), p)		CISESIR	CISELDA	MGSDA	PLDA	MSDA
(0.5, 50)	\(\Delta\)	0.590	0.489	0.690	0.215	0.677
	MSR	0.208	0.086	0.111	0.080	0.100
	TPR	0.960	0.994	0.956	0.998	0.950
	FPR	0.357	0.102	0.090	0.204	0.118
(0.5, 500)	\(\Delta\)	0.438	0.536	0.746	0.241	0.683
	MSR	0.091	0.092	0.122	0.085	0.100
	TPR	0.978	0.990	0.924	1.000	0.942
	FPR	0.030	0.017	0.013	0.076	0.011
(0.5, 1000)	\(\Delta\)	0.322	0.302	0.751	0.218	0.749
	MSR	0.080	0.081	0.119	0.076	0.096
	TPR	0.994	1.000	0.922	1.000	0.900
	FPR	0.019	0.014	0.010	0.023	0.008
(0.9, 50)	\(\Delta\)	0.514	0.602	1.050	0.126	0.833
	MSR	0.070	0.069	0.097	0.067	0.077
	TPR	0.920	0.970	0.660	1.000	0.874
	FPR	0.292	0.082	0.077	0.335	0.059
(0.9, 500)	\(\Delta\)	0.273	0.390	1.048	0.229	0.814
	MSR	0.064	0.069	0.095	0.067	0.073
	TPR	0.988	1.000	0.638	1.000	0.868
	FPR	0.021	0.033	0.009	0.291	0.004
(0.9, 1000)	\(\Delta\)	0.492	0.228	1.047	0.185	0.772
	MSR	0.068	0.065	0.096	0.064	0.066
	TPR	1.000	1.000	0.652	1.000	0.850
	FPR	0.033	0.017	0.003	0.131	0.006

\(\Delta\) is as defined in (10); TPR is the true positive rate; FPR is the false positive rate; MSR is the misclassification rate over a test set of 900 observations. Note again, TPR and FPR are with respect to variable selection. The reported numbers are averages over 50 repetitions

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com