Home

OpenEnsembles is an open source Python toolkit for ensemble clustering. It was designed to allow easy implementation of ensemble clustering, built through combinations of perturbations made to the data or to the algorithms used to cluster the data (for example: by altering the data through transformations on data, selecting different distance metrics, clustering into a different number of clusters (K), or using different clustering algorithms). The many clustering solutions can then be aggregated and visualized and `finished' -- i.e. partitioning the ensemble of solutions into one final solution. Also, you may just be interested in the Python implementations of the many clustering validation metrics that we reimplemented from the R packages. These can be used on individual solutions or final solutions from the ensemble.

For details, see Documentation