class KMeans

This class can be used to do a "kmeans" on a given set of data.

Inheritance:

Public Fields

Sequence* min_cluster

for each example, keep the index of the neirest cluster

Public Methods

KMeans(int n_inputs, int n_gaussians_)

virtual real frameLogProbability(int t, real* inputs)

note that this method returns in fact the euclidean distance between the observation and the neirest cluster

virtual real frameLogProbabilityOneGaussian(int g, real* inputs)

similarly to frameLogProbability, this method returns the euclidean distance between cluster g and the given observation

Public Members

initialize the parameters from the data set to false if you

Inherited from DiagonalGMM:

Public Fields

int n_gaussians

real prior_weights

EMTrainer* initial_kmeans_trainer

MeasurerList* initial_kmeans_trainer_measurers

real* log_weights

real* dlog_weights

real* var_threshold

Sequence* log_probabilities_g

int best_gauss

Sequence* best_gauss_per_frame

real* sum_log_var_plus_n_obs_log_2_pi

real** minus_half_over_var

real** means_acc

Public Methods

virtual void reset()

virtual void setDataSet(DataSet* data_)

virtual void setVarThreshold(real* var_threshold_)

virtual void eMIterInitialize()

virtual void iterInitialize()

virtual real viterbiFrameLogProbability(int t, real* inputs)

virtual void sequenceInitialize(Sequence* inputs)

virtual void eMSequenceInitialize(Sequence* inputs)

virtual void frameEMAccPosteriors(int t, real* inputs, real log_posterior)

virtual void frameViterbiAccPosteriors(int t, real* inputs, real log_posterior)

virtual void eMUpdate()

virtual void update()

virtual void frameBackward(int t, real* f_inputs, real* beta_, real* f_outputs, real* alpha_)

virtual void frameDecision(int t, real* decision)

Inherited from Distribution:

Public Fields

real log_probability

Sequence* log_probabilities

Public Methods

virtual real logProbability(Sequence* inputs)

virtual real viterbiLogProbability(Sequence* inputs)

virtual void eMAccPosteriors(Sequence* inputs, real log_posterior)

virtual void viterbiAccPosteriors(Sequence* inputs, real log_posterior)

virtual void decode(Sequence* inputs)

virtual void eMForward(Sequence* inputs)

virtual void viterbiForward(Sequence* inputs)

virtual void viterbiBackward(Sequence* inputs, Sequence* alpha)

Public Members

Returns the decision of the distribution

Inherited from GradientMachine:

Public Fields

int n_inputs

int n_outputs

Parameters* params

Parameters* der_params

Sequence* beta

Public Methods

virtual void forward(Sequence* inputs)

virtual void backward(Sequence* inputs, Sequence* alpha)

virtual void setPartialBackprop(bool flag=true)

virtual void frameForward(int t, real* f_inputs, real* f_outputs)

virtual void loadXFile(XFile* file)

virtual void saveXFile(XFile* file)

Inherited from Machine:

Public Fields

Sequence* outputs

Inherited from Object:

Public Fields

Allocator* allocator

Public Methods

void addOption(const char* name, int size, void* ptr, const char* help="")

void addIOption(const char* name, int* ptr, int init_value, const char* help="")

void addROption(const char* name, real* ptr, real init_value, const char* help="")

void addBOption(const char* name, bool* ptr, bool init_value, const char* help="")

void addOOption(const char* name, Object** ptr, Object* init_value, const char* help="")

void setOption(const char* name, void* ptr)

void setIOption(const char* name, int option)

void setROption(const char* name, real option)

void setBOption(const char* name, bool option)

void setOOption(const char* name, Object* option)

void load(const char* filename)

void save(const char* filename)

void* operator new(size_t size, Allocator* allocator_=NULL)

void* operator new(size_t size, Allocator* allocator_, void* ptr_)

void operator delete(void* ptr)

Documentation

This class can be used to do a "kmeans" on a given set of data. It has been implemented in the framework of a Distribution that can be trained with EM. This means that the kmeans distance is in fact returned by the method logProbability.

Note that as KMeans is a subclass of DiagonalGMM, they share the same parameter structure. Hence, a DiagonalGMM can be easily initialized by a KMeans.

Sequence* min_cluster

for each example, keep the index of the neirest cluster

initialize the parameters from the data set to false if you

initialize the parameters from the data set to false if you

KMeans(int n_inputs, int n_gaussians_)

virtual real frameLogProbability(int t, real* inputs)

note that this method returns in fact the euclidean distance between the observation and the neirest cluster

virtual real frameLogProbabilityOneGaussian(int g, real* inputs)

similarly to frameLogProbability, this method returns the euclidean distance between cluster g and the given observation

This class has no child classes.

Author:

Samy Bengio (bengio@idiap.ch)

Alphabetic index HTML hierarchy of classes or Java