Conjoint analysis was carried out both for evaluating clusters of consumers in each country and for considering an overall evaluation of all the countries at the same time.
Datta, S. & Datta, S. Methods for evaluating clustering algorithms for gene expression data using a reference set of functional classes.
Similar to a recent study evaluating clustering methods on several biological datasets26, we found that no single clustering is the best performer on all datasets, although certain methods are certainly better than others at retrieving the known modular structure within the data.
The results demonstrate the effectiveness of MCDM methods in evaluating clustering algorithms and indicate that the repeated-bisection method leads to good 2-way clustering solutions on the selected financial risk data sets.
Three variants of the AEA centred on evaluating clustering methods are presented: the baseline AEA, the community-based AEA which groups stations according to journey flows, and the Adaptive AEA which actively modifies clusters to cater for changes in demand.
Evaluating clustering results is very important to the success of clustering tasks.
Their selection of the applied threshold t = 0.65 was driven by evaluating clusters based on available bioactivity data.
It is also worth noting that MC value is capable of evaluating clustering algorithms without the ground truth.
Furthermore, the quality of clustering is characterized by F1-measure and Adjusted Rand Index, which are more suitable for evaluating clustering results than only the true positive rate or accuracy used in [31, 33, 34, 49].
Two measures for evaluating clusters: JaccardC measure and PRC measure, are used that are based on overlaps between the estimated clusters and the known groups of proteins with functional annotations [35].
Our method exploits the known biological functions when evaluating cluster-specific weights in the loss function.
