Unsupervised collaborative boosting of clustering: an unifying framework for multi-view clustering, multiple consensus clusterings and alternative clustering

Jacques-Henri Sublemontier (1)

1 CA

In this paper, we propose a collaborative framework that is able to solve multi-view and alternative clustering problems using some clustering ensemble and semi-supervised clustering principles. We provide a mechanism to control, via a information sharing model, different clustering algorithms to obtain consensus or alternative clustering solutions. The strong point is that our approach does not need to know which clustering algorithms to use nor their parameters

Domains: Neural and Evolutionary Computing [cs.NE]; Machine Learning [cs.LG]; Artificial Intelligence [cs.AI]

Jacques-Henri Sublemontier. Unsupervised collaborative boosting of clustering: an unifying framework for multi-view clustering, multiple consensus clusterings and alternative clustering. International Joint Conference on Neural Networks (IJCNN 2013), IEEE - INNS, Aug 2013, Dallas, United States. ⟨hal-00813201⟩