Semi-supervised co-selection: features and instances by a weighting approach

Feature selection, instance selection and semi-supervised clustering are different challenges for machine learning and data mining communities. While other works have addressed each of these problems separately, in this paper we show how they can be addressed together, simultaneously. We propose an unified framework for semi-supervised co-selection of features and instances, based on weighting constrained clustering. In particular, we define a novel objective function by weighting both instances and features; and constraining the associated partitioning. Experiments are carried out on some known datasets, and results are promising, showing that our proposal outperforms other state-of-the-art algorithms.