Objective: Combine clustering results in order to identify clustering patterns agreed by different numbers of clustering, using frequent closed patterns, and generate new consensus clustering solutions. These consensus clustering solutions are then presented in a tree structure that allows the end-user to better understand the properties of the data space, by identifying strong clusters and unstable instances from the viewpoint of clustering, and determine the best consensus solution according to his/her prior knowledge of the application.

Implementation: R Language.

Collaborators: Atheer AL-NAJDI (Ministry of Higher Education and Scientific Research, Bagdad, Iraq), Nicolas PASQUIER (Lab. I3S, Univ. Côte d'Azur, Nice, France), Frederic PRECIOSO (Lab. I3S, Univ. Côte d'Azur, Nice, France).

Reference publications:
[1] Atheer Al-Najdi, Nicolas Pasquier, Frédéric Precioso. Multiple Consensuses Clustering by Iterative Merging/Splitting of Clustering Patterns in Proceedings of the MLDM'2016 International Conference on Machine Learning and Data Mining in Pattern Recognition, pages 790-804, New York, United States, 16-21 July 2016, LNAI 9729, Springer International Publishing, ISBN 978-3-319-41920-6 (Acceptance Rate: 34%).
[2] Atheer Al-Najdi, Nicolas Pasquier, Frédéric Precioso. Using Frequent Closed Pattern Mining to Solve a Consensus Clustering Problem in Proceedings of the SEKE'2016 International Conference on Software Engineering & Knowledge Engineering, pages 454-461, Redwood City, USA, 1-3 July 2016, KSI Research Inc., ISBN 1-891706-39-X (Acceptance Rate: 29%). SEKE'2016 Third Place Award.
[3] Atheer Al-Najdi, Nicolas Pasquier, Frédéric Precioso. Frequent Closed Patterns Based Multiple Consensus Clustering in Proceedings of the ICAISC'2016 International Conference on Artificial Intelligence and Soft Computing, Part II, pages 14-26, Zakopane, Poland, 12-16 June 2016, LNCS 9693, Springer, ISBN 978-3-319-39384-1.