A0487
Title: Heteroskedastic sparse PCA in high dimensions
Authors: Zhao Ren - University of Pittsburgh (United States) [presenting]
Abstract: Principal component analysis (PCA) is one of the most commonly used dimension reduction and feature extraction techniques. Though it has been well-studied for high-dimensional sparse PCA, little is known when the noise is heteroscedastic, ubiquitous in many scenarios, like biological sequencing data and information network data. An iterative algorithm is proposed for sparse PCA in the presence of heteroskedastic noise, which alternatively updates the estimates of the sparse eigenvectors using the power method with adaptive thresholding in one step. In the other step, it imputes the diagonal values of the sample covariance matrix to reduce the estimation bias due to heteroskedasticity. Our procedure is computationally fast and optimal under the generalized spiked covariance model, assuming the leading eigenvectors are sparse. A comprehensive simulation study demonstrates its robustness and effectiveness in various settings.
