nearestCentroid: Class Label Based on the Minimum Mahalanobis Distance

In FRESA.CAD: Feature Selection Algorithms for Computer Aided Diagnosis

Class Label Based on the Minimum Mahalanobis Distance

Description

The function will return the set of labels of a data set

Usage

   nearestCentroid(dataset,
                  clustermean=NULL,
                  clustercov=NULL, 
                  p.threshold=1.0e-6)

Arguments

dataset	The data set to be labeled
clustermean	The list of cluster centers.
clustercov	The list of cluster covariances
p.threshold	The minimum aceptance p.value

Details

The data set will be labeled based on the nearest cluster label. Points distance with membership probability lower than the acceptance threshold will have the "0" label.

Value

ClusterLabels

The labels of each point

Author(s)

Jose G. Tamez-Pena

FRESA.CAD documentation built on June 26, 2024, 1:07 a.m.