Skip to main content

Table 9 Comparison among DPART, FUM, and FCM in data sets with multiple small subsets

From: Characterization of a Bayesian genetic clustering algorithm based on a Dirichlet process prior and comparison among Bayesian clustering methods

 

Nl = 20

 

Nl = 50

Nl = 10

 

Nl = 20

 

N (10, 10, 200)

N (10, 200, 200)

N (10, 10, 200)

N (50, 50, 200)

N (50, 200, 200)

N (50, 50, 200)

DPART

0.112 (66)

0.032 (92)

0.004 (100)

0.043 (99)

0.040 (99)

0.006 (100)

 

2.88

3.07

3.16

3.05

3.11

3.11

FUM

0.435 (7)

0.012 (99)

0.334 (0)

0.154 (74)

0.035 (100)

0.057 (89)

FCM

0.496 (0)

0.055 (83)

0.364 (17)

0.162 (73)

0.041 (100)

0.072 (86)

  1. The migration rate was 0.003