Semi-supervised Ensemble Clustering Based On Selected Constraint Projection

Semi-supervised ensemble clustering based on selected constraint projection

Abstract

Cluster ensemble study constitutes an important branch of ensemble learning. When compared with traditional single clustering approaches, they are able to integrate multiple clustering results into a unified result which is more accurate and stable. The cluster ensemble approach can be divided into two stages: ensemble generation and consensus function. The objective of the ensemble generation stage is to obtain diverse ensemble members, while the objective of the consensus function is to generate a unified clustering solution as accurately as possible.