2015/02/08 by David B. King, King, David B.
Computer Science · Engineering · Mathematics · #45B05 #45Q05 #60E05 #62H20 #62H25 #62M99 #Control Systems and Identification #FOS: Mathematics #Image and Signal Denoising Methods #Neural Networks and Applications #Statistical Methods and Inference #Statistical and numerical algorithms #Statistics Theory (math.ST)
paper · pdf · doi:10.48550/arxiv.1502.02291
openalex publication_date 2015/02/08 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
In this paper we derive the asymptotic distributions of two distinct\nregularized estimators for functional canonical correlation as well as their\nassociated eigenvalues, eigenvectors and projection operators. The methods we\ndeveloped utilize regularized estimators which approach the functional\noperators based in reproducing kernel Hilbert spaces (RKHS) as the\nregularization parameter approaches zero. In addition to providing some\njustification for the RKHS methods, we explore the asymptotics of regularized\noperators associated with both Tikhinov and truncated singular value\ndecomposition (TSVD) type regularization. Together, these regularization\nmethods represent two of the most commonly utilized forms of regularization.\n