High-dimensional model-based clustering via random projections

Random projections (RPs) have shown to provide promising results in the context of high-dimensional supervised classification. In this work, we address the unsupervised classification issue by exploiting the general idea of RP ensemble. Specifically, we generate a set of low dimensional independent random projections and we perform a model-based clustering on each of them. The top B* projections, i.e. the projections which show the best grouping structure, are then retained. The final partition is obtained by aggregating the chosen classifiers via consensus. The performances of the method are assessed on a set of both real and simulated data.