Abstract:Principal component analysis (PCA) is a widely used tool for data analysis and dimension reduction in applications throughout science and engineering. However, the principal components (PCs) can sometimes be difficult to interpret, because they are linear combinations of all the original variables. To facilitate interpretation, sparse PCA produces modified PCs with sparse loadings, i.e. loadings with very few non-zero elements. In this paper, we propose a new sparse PCA method, namely sparse PCA via regularized SVD (sPCA-rSVD). We use the connection of PCA with singular value decomposition (SVD) of the data matrix and extract the PCs through solving a low rank matrix approximation problem. Regularization penalties are introduced to the corresponding minimization problem to promote sparsity in PC loadings. An efficient iterative algorithm is proposed for computation. Two tuning parameter selection methods are discussed. Some theoretical results are established to justify the use of sPCA-rSVD when only the data covariance matrix is available. In addition, we give a modified definition of variance explained by the sparse PCs. The sPCA-rSVD provides a uniform treatment of both classical multivariate data and high-dimension-low-sample-size (HDLSS) data. Further understanding of sPCA-rSVD and some existing alternatives is gained through simulation studies and real data examples, which suggests that sPCA-rSVD provides competitive results.

Sparse Probabilistic Relational Projection.

Probabilistic Relational PCA.

Sparse Projections over Graph.

Sparse Probabilistic Principal Component Analysis Model for Plant-Wide Process Monitoring

Sparse Local Discriminant Projections for Feature Extraction

Efficient Probabilistic Latent Semantic Analysis with Sparsity Control

Sparse Principal Component Analysis

Sparse principal component analysis via regularized low rank matrix approximation

Sparse Principal Component Analysis via Variable Projection

Structured Principal Component Analysis Model With Variable Correlation Constraint

Sparse Data-Driven Random Projection in Regression for High-Dimensional Data

Sparse Unsupervised Dimensionality Reduction Algorithms

Projection sparse principal component analysis: An efficient least squares method

Perception Preserving Projections

A survey of dimensionality reduction techniques based on random projection

Robust Principal Component Analysis with Projection Learning for Image Classification

Dynamic Principal Subspaces with Sparsity in High Dimensions

Binary Random Projections with Controllable Sparsity Patterns

Sparse Principal Component Analysis via Joint L2,1-Norm Penalty

Sparsity Induced Locality Preserving Projection Approaches for Dimensionality Reduction