Statistics – Machine Learning
Scientific paper
2011-02-20
Statistics
Machine Learning
27 pages, 13 figures
Scientific paper
Research in a number of fields now requires the analysis of "multi-block" data, in which multiple high-dimensional, and fundamentally disparate, datatypes are available for a common set of objects. In this paper we introduce Joint and Individual Variation Explained (JIVE), a general method for the integrated analysis of multi-block data. The method decomposes a multi-block dataset into a sum of three terms: a low-rank approximation capturing joint variation across datatypes, low-rank approximations for structured variation individual to each datatype, and residual noise. This decomposition can be used to quantify the amount of joint variation between datatypes, visually explore joint and individual structure, and reduce the dimensionality of the data. The proposed method represents an extension of Principal Component Analysis (PCA) and has clear advantages over popular two-block methods such as Canonical Correlation Analysis (CCA) and Partial Least Squares (PLS). We apply JIVE to data from The Cancer Genome Atlas (TCGA), where multiple genomic technologies are available for a common set of Glioblastoma Multiforme tumor samples. Software is available at https://genome.unc.edu/jive/.
Hoadley Katherine A.
Lock Eric F.
Marron Stephen J.
Nobel Andrew B.
No associations
LandOfFree
Joint and Individual Variation Explained (JIVE) for Integrated Analysis of Multiple Datatypes does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Joint and Individual Variation Explained (JIVE) for Integrated Analysis of Multiple Datatypes, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Joint and Individual Variation Explained (JIVE) for Integrated Analysis of Multiple Datatypes will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-519690