This repository contains an implementation of multiple sparse canonical correlation analysis regression, as described in the paper. Using this method, multiple data-views can be used to stratify a single data-view (Ing et al. 2019). The code implementing this method is written in MATLAB and is extensively commented. We have also included a MATLAB implementation of the two-view sparse canonical correlation procedure originally proposed by Witten et al (Ing et al. 2019; Witten, Tibshirani, and Hastie 2009; Witten and Tibshirani 2009) (originally released in R).
Ing, Alex, Philipp G. Sämann, Congying Chu, Nicole Tay, Francesca Biondo, Gabriel Robert, Tianye Jia, et al. 2019. “Identification of Neurobehavioural Symptom Groups Based on Shared Brain Mechanisms.” Nature Human Behaviour 3 (12): 1306–18.
Witten, Daniela M., Robert Tibshirani, and Trevor Hastie. 2009. “A Penalized Matrix Decomposition, with Applications to Sparse Principal Components and Canonical Correlation Analysis.” Biostatistics 10 (3): 515–34.
Witten, Daniela M., and Robert J. Tibshirani. 2009. “Extensions of Sparse Canonical Correlation Analysis with Applications to Genomic Data.” Statistical Applications in Genetics and Molecular Biology 8 (June): Article28.