Skip to main content

Table 1 Proportion of well-classified patients according to dist, the distance between the barycenters of the two groups

From: Importance of data structure in comparing two dimension reduction methods for classification of microarray gene expression data

 

α = π/2

α = π/3

α = π/4

α = π/6

α = 0

dist = 1

     

   PLS+DA

0.69(0.05)//2

0.69(0.06)//2

0.64(0.06)//2

0.60(0.06)//2

0.59(0.05)//1

   PCA+DA

0.69(0.04)//3

0.70(0.05)//2

0.66(0.06)//3

0.60(0.05)//3

0.58(0.06)//1

   BGA

0.63(0.05)

0.61(0.06)

0.55(0.06)

0.57(0.05)

0.58(0.05)

dist = 3

     

   PLS+DA

0.93(0.04)//2

0.91(0.03)//2

0.86(0.03)//2

0.71(0.03)//2

0.69(0.06)//1

   PCA+DA

0.94(0.04)//2

0.91(0.03)//2

0.85(0.04)//3

0.71(0.04)//3

0.69(0.05)//2

   BGA

0.90(0.04)

0.79(0.03)

0.73(0.03)

0.70(0.03)

0.67(0.06)

dist = 5

     

   PLS+DA

0.98(0.04)//2

0.97(0.01)//2

0.97(0.02)//2

0.84(0.03)//2

0.79(0.04)//1

   PCA+DA

0.99(0.01)//3

0.98(0.01)//2

0.97(0.02)//2

0.83(0.03)//2

0.79(0.04)//2

   BGA

0.91(0.04)

0.91(0.01)

0.86(0.03)

0.82(0.03)

0.79(0.04)

  1. Mean (Standard deviation)//Median of the optimal number of components over fifty datasets simulated with eccentricity such that ratio = 10. PLS: Partial Least Squares – PCA: Principal Components Analysis – DA: Discriminant Analysis.