Supervised Wavelet Method to Predict Patient Survival from Gene Expression Data

Table 4

Performance of different Cox models for lung cancer dataset (clinical + genomic data).

# Gene	Method	C index ± se	CPE ± se	±  se	LR ± se	IBS ± se
20	Supervised wavelet	0.949 ± 0.006	0.924 ± 0.010	0.669 ± 0.031	72.304 ± 2.589	0.431 ± 0.007
	Supervised PCA	0.907 ± 0.008	0.844 ± 0.009	0.553 ± 0.033	52.020 ± 2.208	0.432 ± 0.007
	Supervised PLS	0.914 ± 0.007	0.849 ± 0.009	0.564 ± 0.035	53.814 ± 2.366	0.435 ± 0.009
15	Supervised wavelet	0.916 ± 0.005	0.855 ± 0.011	0.558 ± 0.031	56.318 ± 3.017	0.433 ± 0.010
	Supervised PCA	0.903 ± 0.007	0.836 ± 0.010	0.540 ± 0.034	53.478 ± 2.585	0.435 ± 0.009
	Supervised PLS	0.908 ± 0.007	0.842 ± 0.012	0.552 ± 0.041	55.526 ± 2.398	0.435 ± 0.006
10	Supervised wavelet	0.906 ± 0.006	0.848 ± 0.008	0.552 ± 0.027	52.746 ± 2.872	0.426 ± 0.006
	Supervised PCA	0.892 ± 0.009	0.831 ± 0.008	0.521 ± 0.029	48.092 ± 2.119	0.426 ± 0.007
	Supervised PLS	0.905 ± 0.009	0.842 ± 0.009	0.542 ± 0.031	51.472 ± 2.562	0.430 ± 0.005
5	Supervised wavelet	0.895 ± 0.008	0.818 ± 0.011	0.499 ± 0.036	51.472 ± 2.760	0.352 ± 0.008
	Supervised PCA	0.883 ± 0.009	0.803 ± 0.010	0.445 ± 0.042	46.336 ± 2.113	0.359 ± 0.008
	Supervised PLS	0.879 ± 0.007	0.814 ± 0.010	0.481 ± 0.029	49.976 ± 2.152	0.355 ± 0.006