latest version v1.9 - last update 24 Nov 2005

lti::centroidClustering Class Reference

Base class for all clustering methods that use centroids or prototypes for describing individual clusters. More...

#include <ltiCentroidClustering.h>

Inheritance diagram for lti::centroidClustering:

[legend]Collaboration diagram for lti::centroidClustering:

[legend]List of all members.

Public Member Functions
	centroidClustering ()
	centroidClustering (const centroidClustering &other)
virtual	~centroidClustering ()
virtual const char *	getTypeName () const
centroidClustering &	copy (const centroidClustering &other)
const parameters &	getParameters () const
virtual bool	classify (const dvector &feature, outputVector &result) const
const dmatrix &	getCentroids () const
virtual bool	train (const dmatrix &input)=0
virtual bool	train (const dmatrix &input, ivector &ids)
virtual bool	write (ioHandler &handler, const bool complete=true) const
virtual bool	read (ioHandler &handler, const bool complete=true)
Protected Attributes
dmatrix	centroids

---

Detailed Description

Base class for all clustering methods that use centroids or prototypes for describing individual clusters.

A classify(const dvector&) method is implemented that should work for most subclasses.

---

Constructor & Destructor Documentation

lti::centroidClustering::centroidClustering (   )

default constructor

lti::centroidClustering::centroidClustering (  const centroidClustering &  other  )

copy constructor Parameters: other  the object to be copied

virtual lti::centroidClustering::~centroidClustering (   )  [virtual]

destructor

---

Member Function Documentation

virtual bool lti::centroidClustering::classify (  const dvector &  feature, outputVector &  result )  const [virtual]

classifies a new data point. This method can be used by all centroid classifiers if they properly initialize the classifier::output member. I.e. for each output unit the list of probabilities may contain only one entry with probability one and each id may be used only once. Since the classifier::output expects a probability vector as result from the clustering method, the dissimilarity measure usually returned must be converted to a similarity measure. Here, this is done by first normalizing the vector of dissimilarities, then subtracting this vector from a vector filled with ones and, finally, normalizing the result. This operation yields similarity values between zero and one for arbitrary scales of the feature space. Information about possible reject cases is lost through the first normalizaton, though. Parameters: feature  vector to be classified result  result as described above Returns: true if successful, false otherwise Implements lti::unsupervisedClassifier.

centroidClustering& lti::centroidClustering::copy (  const centroidClustering &  other  )

copy data of "other" functor. Parameters: other  the functor to be copied Returns: a reference to this functor object

const dmatrix& lti::centroidClustering::getCentroids (   )  const

Returns a const reference to the centroids of the clustering. Returns: const reference to the centroids

const parameters& lti::centroidClustering::getParameters (   )  const

returns used parameters Reimplemented from lti::clustering. Reimplemented in lti::adaptiveKMeans, lti::competitiveAgglomeration, lti::fuzzyCMeans, and lti::kMeansClustering.

virtual const char* lti::centroidClustering::getTypeName (   )  const [virtual]

returns the name of this type ("centroidClustering") Reimplemented from lti::clustering. Reimplemented in lti::adaptiveKMeans, lti::competitiveAgglomeration, lti::fuzzyCMeans, and lti::kMeansClustering.

virtual bool lti::centroidClustering::read (  ioHandler &  handler, const bool  complete = true )  [virtual]

read the rbf classifier from the given ioHandler Parameters: handler  the ioHandler to be used complete  if true (the default) the enclosing begin/end will be also written, otherwise only the data block will be written. Returns: true if write was successful Reimplemented from lti::classifier.

virtual bool lti::centroidClustering::train (  const dmatrix &  input, ivector &  ids )  [virtual]

Calls clustering::train(const dmatrix&, ivector&). Reimplemented from lti::clustering. Reimplemented in lti::adaptiveKMeans, lti::competitiveAgglomeration, lti::fuzzyCMeans, and lti::kMeansClustering.

virtual bool lti::centroidClustering::train (  const dmatrix &  input  )  [pure virtual]

Declared so it wont be forgotten. Implements lti::clustering. Implemented in lti::adaptiveKMeans, lti::competitiveAgglomeration, lti::fuzzyCMeans, and lti::kMeansClustering.

virtual bool lti::centroidClustering::write (  ioHandler &  handler, const bool  complete = true )  const [virtual]

write the rbf classifier in the given ioHandler Parameters: handler  the ioHandler to be used complete  if true (the default) the enclosing begin/end will be also written, otherwise only the data block will be written. Returns: true if write was successful Reimplemented from lti::classifier.

---

Member Data Documentation

dmatrix lti::centroidClustering::centroids [protected]

matrix containing the centroids formed by the clustering method

---

The documentation for this class was generated from the following file:

• ltiCentroidClustering.h

---