Cluster ensembles --- a knowledge reuse framework for combining multiple partitions
The Journal of Machine Learning Research
Combining Multiple Clusterings Using Evidence Accumulation
IEEE Transactions on Pattern Analysis and Machine Intelligence
The Dissimilarity Representation for Pattern Recognition: Foundations And Applications (Machine Perception and Artificial Intelligence)
Weighted Cluster Ensemble Using a Kernel Consensus Function
CIARP '08 Proceedings of the 13th Iberoamerican congress on Pattern Recognition: Progress in Pattern Recognition, Image Analysis and Applications
Weighted association based methods for the combination of heterogeneous partitions
Pattern Recognition Letters
Weighted ensemble of algorithms for complex data clustering
Pattern Recognition Letters
Hi-index | 0.00 |
Cluster ensemble is a promising technique for improving the clustering results. An alternative to generate the cluster ensemble is to use different representations of the data and different similarity measures between objects. This way, it is produced a cluster ensemble conformed by heterogeneous partitions obtained with different point of views of the faced problem. This diversity enhances the cluster ensemble but, it restricts the combination process since it makes difficult the use of the original data. In this paper, in order to solve these limitations, we propose a unified representation of the objects taking into account the whole information in the cluster ensemble. This representation allows working with the original data of the problem regardless of the used generation mechanism. Also, this new representation is embedded in the WKF [1] algorithm making a more robust cluster ensemble method. Experimental results with numerical, categorical and mixed datasets show the accuracy of the proposed method.