The nature of statistical learning theory
The nature of statistical learning theory
IEEE Transactions on Pattern Analysis and Machine Intelligence
Cluster ensembles --- a knowledge reuse framework for combining multiple partitions
The Journal of Machine Learning Research
Clustering Ensembles: Models of Consensus and Weak Partitions
IEEE Transactions on Pattern Analysis and Machine Intelligence
Cumulative Voting Consensus Method for Partitions with Variable Number of Clusters
IEEE Transactions on Pattern Analysis and Machine Intelligence
Collaborative clustering with background knowledge
Data & Knowledge Engineering
Semi-supervised classification method for dynamic applications
Fuzzy Sets and Systems
Pattern Recognition
Weighted partition consensus via kernels
Pattern Recognition
Ensemble clustering in the belief functions framework
International Journal of Approximate Reasoning
Combining multiple clusterings using similarity graph
Pattern Recognition
Hi-index | 0.00 |
Assessing the number of clusters of statistical populations is a challenging problem in unsupervised learning. In this paper, we propose to overcome this problem by estimating the number of clusters using a novel clustering ensemble scheme. This one combines clustering and classification methods in order to increase the clustering performances. In the first time, the proposed approach divides the patterns into stable and ambiguous sets. The stable set gathers the patterns belonging to one cluster while the ambiguous set corresponds to ambiguous patterns located between different clusters. To detect the appropriate number of clusters, the proposed approach ignores ambiguous patterns and preserves the stable set as good "prototypes". Then, the different partitions obtained from the stable set are evaluated by several cluster validation criteria. Finally, the patterns of the unstable set are assigned to the obtained clusters by supervised classifier.