Novel soft subspace clustering with multi-objective evolutionary approach for high-dimensional data

Many conventional soft subspace clustering techniques merge several criteria into a single objective to improve performance; however, the weighting parameters become important but difficult to set. In this paper, a novel soft subspace clustering with a multi-objective evolutionary approach (MOEASSC) is proposed to this problem. This clustering method considers two types of criteria as multiple objectives and optimizes them simultaneously by using a modified multi-objective evolutionary algorithm with new encoding and operators. An indicator called projection similarity validity index (PSVIndex) is designed to select the best solution and cluster number. Experiments on many datasets demonstrate the usefulness of MOEASSC and PSVIndex, and show that our algorithm is insensitive to its parameters and is scalable to large datasets.

► A soft subspace clustering with multi-objective evolutionary approach is proposed. ► A multi-objective evolutionary approach is designed with several characteristics. ► An index is designed to identify the best solution and the cluster number. ► The experimental results are presented on kinds of datasets. ► The algorithm is insensitive to its parameters and scalable to large dataset.