Skip to main content

Table 3 Cluster accuracy and stability on the completely synthetic data with four repeated measurements at high noise level

From: Clustering gene-expression data with repeated measurements

(a) Cluster accuracy*
Algorithm Similarity measure/model Average SD-weighted CV-weighted FITSS IMM
Average linkage Elliptical NA NA NA NA 0.911 (0.122)
Complete linkage Elliptical NA NA NA NA 0.910 (0.123)
k-means Distance 0.326 (0.136) 0.801 (0.037) 0.666 (0.098) NA NA
Complete linkage Distance 0.498 (0.113) 0.798 (0.144) 0.660 (0.098) 0.014 (0.030) NA
Centroid linkage Distance 0.000 (0.000) 0.762 (0.113) 0.315 (0.156) 0.000 (0.000) NA
Average linkage Distance 0.000 (0.000) 0.713 (0.217) 0.256 (0.071) 0.000 (0.000) NA
MCLUST-HC NA 0.608 (0.173) NA NA 0.480 (0.052) NA
Average linkage Spherical NA NA NA NA 0.589 (0.212)
Complete linkage Spherical NA NA NA NA 0.559 (0.358)
k-means Correlation 0.556 (0.121) 0.499 (0.130) 0.394 (0.194) NA NA
Average linkage Correlation 0.389 (0.151) 0.519 (0.159) 0.378 (0.081) 0.291 (0.130) NA
Complete linkage Correlation 0.450 (0.122) 0.518 (0.159) 0.484 (0.156) 0.341 (0.112) NA
Centroid linkage Correlation 0.358 (0.097) 0.261 (0.101) 0.215 (0.096) 0.358 (0.097) NA
(b) Cluster stability
Algorithm Similarity measure/model Average SD-weighted CV-weighted FITSS IMM
Average linkage Elliptical NA NA NA NA 0.948 (0.099)
Average linkage Distance 0.208 (0.075) 0.932 (0.049) 0.812 (0.138) 0.381 (0.103) NA
Centroid linkage Distance 0.211 (0.087) 0.920 (0.097) 0.779 (0.169) 0.211 (0.087) NA
Complete linkage Elliptical NA NA NA NA 0.912 (0.113)
k-means Distance 0.508 (0.153) 0.882 (0.165) 0.686 (0.147) NA NA
Average linkage Correlation Correlation 0.721 (0.060) 0.782 (0.083) 0.692 (0.103) 0.855 (0.015) NA
Complete linkage Distance 0.429 (0.105) 0.803 (0.159) 0.582 (0.083) 0.126 (0.025) NA
Centroid linkage Correlation 0.731 (0.130) 0.497 (0.109) 0.430 (0.099) 0.731 (0.130) NA
k-means Correlation 0.719 (0.070) 0.515 (0.046) 0.382 (0.132) NA NA
Average linkage Spherical NA NA NA NA 0.674 (0.094)
MCLUST-HC NA 0.584 (0.093) NA NA 0.527 (0.026) NA
Complete linkage Correlation 0.580 (0.044) 0.497 (0.094) 0.493 (0.057) 0.353 (0.046) NA
Complete linkage Spherical NA NA NA NA 0.472 (0.288)
  1. *Each entry shows the average adjusted Rand index of the corresponding clustering approach with the six classes. We ran our experiments on five randomly generated synthetic datasets, and show the average results with the standard deviation of the adjusted Rand index in brackets. A high average adjusted Rand index represents close agreement with the classes on average. Each entry shows the average adjusted Rand index of the original clustering result with clusters from remeasured data. The standard deviation of the average adjusted Rand indices over five sets of randomly generated synthetic remeasured data is shown in brackets. A high average adjusted Rand index means that clusters from remeasured data are in close agreement with clusters from the original data. The external knowledge is not used in evaluating cluster stability. In both parts of the table, the maximum average adjusted Rand index of each row is shown in bold. The algorithms (rows) are sorted in descending order of the maximum average adjusted Rand index in each row.