Sampling properties of color Independent Component Analysis

JOURNAL OF MULTIVARIATE ANALYSIS

卷:181

Article

Lee, Seonjoo^1,2,3  Shen, Haipeng⁴  Young Truong⁵ 

[1] Columbia Univ, Dept Psychiat & Biostat, New York, NY 10027 USA

[2] New York State Psychiat Inst & Hosp, Mental Hlth Data Sci, New York, NY 10032 USA

[3] Res Fdn Mental Hyg Inc, New York, NY USA

[4] Univ Hong Kong, Fac Business & Econ, Innovat & Informat Management, Hong Kong, Peoples R China

[5] Univ N Carolina, Dept Biostat, Chapel Hill, NC 27515 USA

关键词: Blind source separation;    Multivariate analysis;    Spectral density estimation;    Time series;    Whittle likelihood;   

DOI  :  10.1016/j.jmva.2020.104692

来源: Elsevier

【 摘 要 】

Independent Component Analysis (ICA) offers an effective data-driven approach for blind source extraction encountered in many signal and image processing problems. Although many ICA methods have been developed, they have received relatively little attention in the statistics literature, especially in terms of rigorous theoretical investigation for statistical inference. The current paper aims at narrowing this gap and investigates the statistical sampling properties of the colorICA (cICA) method. The cICA incorporates the correlation structure within sources through parametric time series models in the frequency domain and outperforms several existing ICA alternatives numerically. We establish the consistency and asymptotic normality of the cICA estimates, which then enables statistical inference based on the estimates. These asymptotic properties are further validated using simulation studies. (C) 2020 Elsevier Inc. All rights reserved.

【 授权许可】

Free