Clustering scores
Clustering scoresCramer_V:
Rand Index: the index produces a result in the range [0,1], where a value of 1.0 indicates that the labels and the calculated clusters are identical. A high value for this measure generally indicates a high level of agreement between two clustering.
Mutual information: Criterion can be used as an external measure for clustering (Strehl et al., 2000). When no assumption is made on the cluster structure, the index can be used to compare clustering algorithms such as k-medoids.
Homogeneity: each cluster contains only members of a single class.
Completeness: all members of a given class are assigned to the same cluster.
Their harmonic mean called V measure.
clustering_1: Connectivity ROI Texture ( input )
clustering_2: Connectivity ROI Texture ( input )
time_step_max: Integer ( input )The number of parcels in the cortical parcellations ()
output_dir: Directory ( output )The output directory.
ybound: ListOf( Float ) ( optional, input )The boundaries of the scale Y (for eaxample: 0.2 0.8).
ignore_Kopt2: Boolean ( input )If this parameter is checked and the kopt is 2, the optimal number is the second largest kopt.
Toolbox : Constellation
User level : 2
Identifier :
clustering_scoresFile name :
brainvisa/toolboxes/constellation/processes/clustering_evaluation/cluster_validity/clustering_scores.pySupported file formats :
clustering_1 :GIFTI file, GIFTI file, Textureclustering_2 :GIFTI file, GIFTI file, Textureoutput_dir :Directory, Directory
