Principal Components Analysis of Scalar, Vector, and Mesh Vertex Data
An implementation of standard PCA algorithms for use on
scalar or vector data sets. Kernel PCA is implemented in this class as well, where the data sets are scalar or
vector valued functions assigned at each of the points in a PointSet. A Gaussian Distance Kernel class is
provided with the PCA class.
This contribution is part of a shape analysis software pipeline created at Johns Hopkins. PCA will be used
to develop a set of basis vectors for use with Gaussian Random Field analysis. The output of PCA will be
analyzed for significance with various statistical methods such as t-tests built upon the built-in statistical
capabilities of ITK.
scalar or vector data sets. Kernel PCA is implemented in this class as well, where the data sets are scalar or
vector valued functions assigned at each of the points in a PointSet. A Gaussian Distance Kernel class is
provided with the PCA class.
This contribution is part of a shape analysis software pipeline created at Johns Hopkins. PCA will be used
to develop a set of basis vectors for use with Gaussian Random Field analysis. The output of PCA will be
analyzed for significance with various statistical methods such as t-tests built upon the built-in statistical
capabilities of ITK.