Skip to main content

Table 1 Feature selection through singular vectors (SVs) in data 2

From: Sparse kernel canonical correlation analysis for discovery of nonlinear interactions in high-dimensional data

  1st SV (η (1)) 2nd SV (η (2)) 3rd SV (η (3))
η 1 0.98 (0.002) 0.00 (0.018) 0.00 (0.001)
η 2 0.00 (0.003) 0.21 (0.033) 0.00 (0.001)
η 3 0.00 (0.001) 0.00 (0.010) 0.22 (0.029)
η 4 0.22 (0.013) 0.00 (0.017) 0.00 (0.005)
η 5 0.00 (0.000) 0.98 (0.004) 0.00 (0.005)
η 6 0.00 (0.004) 0.00 (0.002) 0.98 (0.003)
  1st SV (μ (1)) 2nd SV (μ (2)) 3rd SV (μ (3))
μ 1 0.99 (0.005) 0.01 (0.022) 0.01 (0.014)
μ 2 0.01 (0.027) 0.99 (0.004) 0.01 (0.015)
μ 3 0.01 (0.024) 0.01 (0.018) 0.99 (0.003)
μ 4 0.01 (0.023) 0.01 (0.026) 0.01 (0.017)
  1. These results show mean weight coefficients (standard deviation) in 100 simulation runs. Significant weight coefficients are bold faced