PDF version of this entire document

Methods

A t the core of our methodology ought to exist a mechanism which removes expressions-imposed anatomical variance, as well as complexity such as expression representation. We address these issues with aim of acquiring serialised models. When models are built it ought to be possible to use discrepancy as a measure of dissimilarity and test systematically on a large set, then plug everything into a G-PCA/GMDS algorithm and provide direct comparisons. Here are some key components that make up the method.



Subsections

Roy Schestowitz 2012-01-08