Clustering of multivariate binary data with dimension reduction via L1-regularized likelihood maximization

作者：

Highlights：

• We propose a novel clustering method of multivariate binary data.

• The proposed method provides a low-dimensional representation of clusters.

• The proposed method overcomes the conventional tandem analysis.

• The proposed method conducts feature selection for clustering.

摘要

Highlights•We propose a novel clustering method of multivariate binary data.•The proposed method provides a low-dimensional representation of clusters.•The proposed method overcomes the conventional tandem analysis.•The proposed method conducts feature selection for clustering.

论文关键词：Binary data,Clustering,Dimension reduction,EM algorithm,Latent class analysis,Sparsity

论文评审过程：Received 25 June 2014, Revised 27 May 2015, Accepted 29 May 2015, Available online 9 June 2015, Version of Record 19 August 2015.

论文官网地址：https://doi.org/10.1016/j.patcog.2015.05.026