### Table 3: Analysis of SwissProt keywords composition in clusters calculated for the SwissProt database

2005

### Table 1. Comparison between the binary tree and 4-ary clustering algorithms using the MIPS complexes database. See text for complete details and analysis.

2003

"... In PAGE 13: ... For each complex family, and each clustering algorithm (binary and 4-ary), we determined the cluster (internal node in the tree) that holds at least half of the genes in that family, and has the highest ratio of genes from that family to the total number of genes in that cluster. We report the results for the 8 families having more than 30 genes in Table1 . For three families (Proteasome, Respiration chain, and Translation) our 4-ary tree contains clusters in which these families are significantly overrepre- sented.... ..."

Cited by 19

### Table 1. Comparison between the binary tree and 4-ary clustering algorithms using the MIPS complexes database. See text for complete details and analysis.

2003

Cited by 19

### Table 1. Comparison between the binary tree and 4-ary clustering algorithms using the MIPS complexes database. See text for complete details and analysis.

2003

Cited by 19

### Table 5.5. Summary of clustering results for FeatureSet2 (after feature selection and with PM latency and gaze shift features added)

2007

### Table 2 Summary of feature-based time series clustering algorithms

2005

"... In PAGE 11: ... The clustering error is computed at the end of each level as the sum of number of incorrectly clustered objects for each cluster divided by the cardinality of the dataset. Table2 summarizes major components used in each feature-based clustering algorithm. They all can handle series with unequal length because the feature extraction operation takes care of the issue.... ..."

### Table 15: Residual correlations between series in each cluster

"... In PAGE 16: ... The standardised first difference of the logarithm of each the series excluding the March 1998 value, in each of the clusters (NSW, NT), (QLD, SA, WA), (VIC, ACT), were pooled together and fitted with the relevant AR (k) model. The residual correlations between the series in each cluster is given in Table15 , from where it is clear that the series are related. lt;INSERT TABLE 15 HERE gt; The forecasts from the pooled and individual models, percentage decrease in the mean square forecast error between and individual and pooled models of each of the series as well as the percentage decrease in group average mean square forecast error between the pooled and individual models for the one-step-ahead forecasts, (that is for March 1998) are given in Table 16.... ..."

### Table 4: Results of clustering in databases with known classes

2003

"... In PAGE 18: ...The results presented in Table4 show that Algorithm 2.1 gives better results for all datasets, except the vehicles dataset where the results are almost similar.... ..."

Cited by 2