WeightMatrixByGrid

MLModule

genre	FiberClustering
authors	Jan Klein, Philip Bittihn, Peter Ledochowitsch
package	FMEwork/Release
dll	MLFiberClustering
definition	MLFiberClustering.def
keywords	DTI, fiber, clustering, grid, affinity, matrix

Purpose

The module WeightMatrixByGrid calculates an affinity matrix from a fiber set using the fiber-grid method proposed in the paper Grid-Based Spectral Fiber Clustering.

Details

There are two natural ways to define an affinity matrix:

1. weight matrix: A high value of the element ij corresponds to a high affinity between i and j.
2. distance matrix: A high value of the element ij corresponds to a low affinity between i and j.

Each object of the class Matrix has a tag indicating the type of the matrix (e.g. distanceMatrix or weightMatrix).

The module ClusteringAlgorithms checks those tags and converts the type of a matrix if necessary.

Windows

Default Panel

Input Fields

inputFiberSet

name: inputFiberSet, type: MLBase

A FiberSet.

Output Fields

outMatrix

name: outMatrix, type: MLBase

Affinity matrix as a normalized weight matrix.

outCurve

name: outCurve, type: MLBase