A Framework for Semi-supervised Clustering Based on Dimensionality Reduction

In machine learning and pattern recognition fields, collecting labeled examples costs human efforts, while vast amounts of unlabelled data are often readily available and offer some additional information. In practice, many applications require a dimensionality reduction method to deal with the partially labeled problem. In this paper, we propose a semi-supervised clustering framework, which is based on feature projection and semi-supervised fuzzy clustering. High dimensional data is mapped into a low dimensional space with feature projection. A mapping of learning data is generalized to a new datum which is classified in the low dimensional space with constrained fuzzy clustering. With the experiment on different datasets, the results show the method has good clustering performance for handling data of high dimensionality.