SHOGUN  4.2.0
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Modules Pages
List of all members | Public Member Functions | Public Attributes | Protected Member Functions | Static Protected Member Functions | Protected Attributes
CSingleFITCInference Class Referenceabstract

Detailed Description

The Fully Independent Conditional Training inference base class for Laplace and regression for 1-D labels (1D regression and binary classification)

This base class implements the (explicit) derivatives of negative log marginal likelihood wrt hyperparameter for FITC regression and FITC single Laplace. For FITC single Laplace, we can compute further implicit derivatives. For FITC regression, these explicit derivatives are the full derivatives.

For more details, see QuiƱonero-Candela, Joaquin, and Carl Edward Rasmussen. "A unifying view of sparse approximate Gaussian process regression." The Journal of Machine Learning Research 6 (2005): 1939-1959.

Note that the number of inducing points (m) is usually far less than the number of input points (n). (the time complexity is computed based on the assumption m < n)

This specific implementation was inspired by the infFITC.m and infFITC_Laplace.m file in the GPML toolbox.

Warning: the time complexity of method, CSingleFITCInference::get_derivative_wrt_kernel(const TParameter* param), depends on the implementation of virtual kernel method, CKernel::get_parameter_gradient_diagonal(param, i). The default time complexity of the kernel method can be O(n^2)

Definition at line 68 of file SingleFITCInference.h.

Inheritance diagram for CSingleFITCInference:
[legend]

Public Member Functions

 CSingleFITCInference ()
 
 CSingleFITCInference (CKernel *kernel, CFeatures *features, CMeanFunction *mean, CLabels *labels, CLikelihoodModel *model, CFeatures *inducing_features)
 
virtual ~CSingleFITCInference ()
 
virtual const char * get_name () const
 
virtual void set_kernel (CKernel *kern)
 
virtual void optimize_inducing_features ()
 
virtual void set_lower_bound_of_inducing_features (SGVector< float64_t > bound)
 
virtual void set_upper_bound_of_inducing_features (SGVector< float64_t > bound)
 
virtual void set_tolearance_for_inducing_features (float64_t tol)
 
virtual void set_max_iterations_for_inducing_features (int32_t it)
 
virtual void enable_optimizing_inducing_features (bool is_optmization, FirstOrderMinimizer *minimizer=NULL)
 
virtual EInferenceType get_inference_type () const
 
virtual void set_inducing_features (CFeatures *feat)
 
virtual CFeaturesget_inducing_features ()
 
virtual SGVector< float64_tget_alpha ()
 
virtual SGMatrix< float64_tget_cholesky ()
 
virtual void update ()=0
 
virtual void set_inducing_noise (float64_t noise)
 
virtual float64_t get_inducing_noise ()
 
virtual SGVector< float64_tget_posterior_mean ()=0
 
virtual SGMatrix< float64_tget_posterior_covariance ()=0
 
virtual float64_t get_negative_log_marginal_likelihood ()=0
 
float64_t get_marginal_likelihood_estimate (int32_t num_importance_samples=1, float64_t ridge_size=1e-15)
 
virtual CMap< TParameter
*, SGVector< float64_t > > * 
get_negative_log_marginal_likelihood_derivatives (CMap< TParameter *, CSGObject * > *parameters)
 
virtual SGVector< float64_tget_diagonal_vector ()=0
 
virtual CMap< TParameter
*, SGVector< float64_t > > * 
get_gradient (CMap< TParameter *, CSGObject * > *parameters)
 
virtual SGVector< float64_tget_value ()
 
virtual CFeaturesget_features ()
 
virtual void set_features (CFeatures *feat)
 
virtual CKernelget_kernel ()
 
virtual CMeanFunctionget_mean ()
 
virtual void set_mean (CMeanFunction *m)
 
virtual CLabelsget_labels ()
 
virtual void set_labels (CLabels *lab)
 
CLikelihoodModelget_model ()
 
virtual void set_model (CLikelihoodModel *mod)
 
virtual float64_t get_scale () const
 
virtual void set_scale (float64_t scale)
 
virtual bool supports_regression () const
 
virtual bool supports_binary () const
 
virtual bool supports_multiclass () const
 
virtual SGMatrix< float64_tget_multiclass_E ()
 
virtual void register_minimizer (Minimizer *minimizer)
 
virtual CSGObjectshallow_copy () const
 
virtual CSGObjectdeep_copy () const
 
virtual bool is_generic (EPrimitiveType *generic) const
 
template<class T >
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
template<>
void set_generic ()
 
void unset_generic ()
 
virtual void print_serializable (const char *prefix="")
 
virtual bool save_serializable (CSerializableFile *file, const char *prefix="")
 
virtual bool load_serializable (CSerializableFile *file, const char *prefix="")
 
void set_global_io (SGIO *io)
 
SGIOget_global_io ()
 
void set_global_parallel (Parallel *parallel)
 
Parallelget_global_parallel ()
 
void set_global_version (Version *version)
 
Versionget_global_version ()
 
SGStringList< char > get_modelsel_names ()
 
void print_modsel_params ()
 
char * get_modsel_param_descr (const char *param_name)
 
index_t get_modsel_param_index (const char *param_name)
 
void build_gradient_parameter_dictionary (CMap< TParameter *, CSGObject * > *dict)
 
bool has (const std::string &name) const
 
template<typename T >
bool has (const Tag< T > &tag) const
 
template<typename T , typename U = void>
bool has (const std::string &name) const
 
template<typename T >
void set (const Tag< T > &_tag, const T &value)
 
template<typename T , typename U = void>
void set (const std::string &name, const T &value)
 
template<typename T >
get (const Tag< T > &_tag) const
 
template<typename T , typename U = void>
get (const std::string &name) const
 
virtual void update_parameter_hash ()
 
virtual bool parameter_hash_changed ()
 
virtual bool equals (CSGObject *other, float64_t accuracy=0.0, bool tolerant=false)
 
virtual CSGObjectclone ()
 

Public Attributes

SGIOio
 
Parallelparallel
 
Versionversion
 
Parameterm_parameters
 
Parameterm_model_selection_parameters
 
Parameterm_gradient_parameters
 
uint32_t m_hash
 

Protected Member Functions

virtual float64_t get_derivative_related_cov (SGVector< float64_t > ddiagKi, SGMatrix< float64_t > dKuui, SGMatrix< float64_t > dKui)
 
virtual float64_t get_derivative_related_cov (SGVector< float64_t > ddiagKi, SGMatrix< float64_t > dKuui, SGMatrix< float64_t > dKui, SGVector< float64_t > v, SGMatrix< float64_t > R)
 
virtual float64_t get_derivative_related_cov_helper (SGMatrix< float64_t > dKuui, SGVector< float64_t > v, SGMatrix< float64_t > R)
 
virtual SGVector< float64_tget_derivative_wrt_inducing_noise (const TParameter *param)
 
virtual SGVector< float64_tget_derivative_related_cov_diagonal ()
 
virtual float64_t get_derivative_related_mean (SGVector< float64_t > dmu)
 
virtual SGVector< float64_tget_derivative_related_inducing_features (SGMatrix< float64_t > BdK, const TParameter *param)
 
virtual void update_alpha ()=0
 
virtual void update_chol ()=0
 
virtual void update_deriv ()=0
 
virtual SGVector< float64_tget_derivative_wrt_likelihood_model (const TParameter *param)=0
 
virtual SGVector< float64_tget_derivative_wrt_mean (const TParameter *param)
 
virtual SGVector< float64_tget_derivative_wrt_inducing_features (const TParameter *param)
 
virtual SGVector< float64_tget_derivative_wrt_inference_method (const TParameter *param)
 
virtual SGVector< float64_tget_derivative_wrt_kernel (const TParameter *param)
 
virtual void check_bound (SGVector< float64_t > bound, const char *name)
 
virtual void check_fully_sparse ()
 
virtual void convert_features ()
 
virtual void check_features ()
 
virtual void check_members () const
 
virtual void update_train_kernel ()
 
virtual void compute_gradient ()
 
virtual void load_serializable_pre () throw (ShogunException)
 
virtual void load_serializable_post () throw (ShogunException)
 
virtual void save_serializable_pre () throw (ShogunException)
 
virtual void save_serializable_post () throw (ShogunException)
 
template<typename T >
void register_param (Tag< T > &_tag, const T &value)
 
template<typename T >
void register_param (const std::string &name, const T &value)
 

Static Protected Member Functions

static void * get_derivative_helper (void *p)
 

Protected Attributes

SGVector< float64_tm_al
 
SGVector< float64_tm_t
 
SGMatrix< float64_tm_B
 
SGVector< float64_tm_w
 
SGMatrix< float64_tm_Rvdd
 
SGMatrix< float64_tm_V
 
SGVector< float64_tm_lower_bound
 
SGVector< float64_tm_upper_bound
 
float64_t m_max_ind_iterations
 
float64_t m_ind_tolerance
 
bool m_opt_inducing_features
 
bool m_fully_sparse
 
CLockm_lock
 
FirstOrderMinimizerm_inducing_minimizer
 
SGMatrix< float64_tm_inducing_features
 
float64_t m_log_ind_noise
 
SGMatrix< float64_tm_kuu
 
SGMatrix< float64_tm_ktru
 
SGMatrix< float64_tm_Sigma
 
SGVector< float64_tm_mu
 
SGVector< float64_tm_ktrtr_diag
 
Minimizerm_minimizer
 
CKernelm_kernel
 
CMeanFunctionm_mean
 
CLikelihoodModelm_model
 
CFeaturesm_features
 
CLabelsm_labels
 
SGVector< float64_tm_alpha
 
SGMatrix< float64_tm_L
 
float64_t m_log_scale
 
SGMatrix< float64_tm_ktrtr
 
SGMatrix< float64_tm_E
 
bool m_gradient_update
 

Constructor & Destructor Documentation

default constructor

Definition at line 42 of file SingleFITCInference.cpp.

CSingleFITCInference ( CKernel kernel,
CFeatures features,
CMeanFunction mean,
CLabels labels,
CLikelihoodModel model,
CFeatures inducing_features 
)

constructor

Parameters
kernelcovariance function
featuresfeatures to use in inference
meanmean function
labelslabels of the features
modellikelihood model to use
inducing_featuresfeatures to use

Definition at line 47 of file SingleFITCInference.cpp.

~CSingleFITCInference ( )
virtual

Definition at line 64 of file SingleFITCInference.cpp.

Member Function Documentation

void build_gradient_parameter_dictionary ( CMap< TParameter *, CSGObject * > *  dict)
inherited

Builds a dictionary of all parameters in SGObject as well of those of SGObjects that are parameters of this object. Dictionary maps parameters to the objects that own them.

Parameters
dictdictionary of parameters to be built.

Definition at line 630 of file SGObject.cpp.

void check_bound ( SGVector< float64_t bound,
const char *  name 
)
protectedvirtualinherited

check the bound constraint is vailid or not

Parameters
boundbound constrains of inducing features
namethe name of the bound

Definition at line 282 of file SingleSparseInference.cpp.

void check_features ( )
protectedvirtualinherited

check whether features and inducing features are set

Definition at line 48 of file SparseInference.cpp.

void check_fully_sparse ( )
protectedvirtualinherited

check whether the provided kernel can compute the gradient wrt inducing features

Note that currently we check the name of the provided kernel to determine whether the kernel can compute the derivatives wrt inducing_features

The name of a supported Kernel must end with "SparseKernel"

Definition at line 176 of file SingleSparseInference.cpp.

void check_members ( ) const
protectedvirtualinherited

check if members of object are valid for inference

Reimplemented from CInference.

Reimplemented in CFITCInferenceMethod, and CVarDTCInferenceMethod.

Definition at line 127 of file SparseInference.cpp.

CSGObject * clone ( )
virtualinherited

Creates a clone of the current object. This is done via recursively traversing all parameters, which corresponds to a deep copy. Calling equals on the cloned object always returns true although none of the memory of both objects overlaps.

Returns
an identical copy of the given object, which is disjoint in memory. NULL if the clone fails. Note that the returned object is SG_REF'ed

Definition at line 747 of file SGObject.cpp.

void compute_gradient ( )
protectedvirtualinherited
void convert_features ( )
protectedvirtualinherited

convert inducing features and features to the same represention

Note that these two kinds of features can be different types. The reasons are listed below.

  1. The type of the gradient wrt inducing features is float64_t, which is used to update inducing features
  2. Reason 1 implies that the type of inducing features can be float64_t while the type of features does not required as float64_t
  3. Reason 2 implies that the type of features must be a subclass of CDotFeatures, which can represent features as float64_t

Definition at line 53 of file SparseInference.cpp.

CSGObject * deep_copy ( ) const
virtualinherited

A deep copy. All the instance variables will also be copied.

Definition at line 231 of file SGObject.cpp.

void enable_optimizing_inducing_features ( bool  is_optmization,
FirstOrderMinimizer minimizer = NULL 
)
virtualinherited

whether enable to opitmize inducing features

Parameters
is_optmizationenable optimization
minimizerminimizer used in optimization

Definition at line 321 of file SingleSparseInference.cpp.

bool equals ( CSGObject other,
float64_t  accuracy = 0.0,
bool  tolerant = false 
)
virtualinherited

Recursively compares the current SGObject to another one. Compares all registered numerical parameters, recursion upon complex (SGObject) parameters. Does not compare pointers!

May be overwritten but please do with care! Should not be necessary in most cases.

Parameters
otherobject to compare with
accuracyaccuracy to use for comparison (optional)
tolerantallows linient check on float equality (within accuracy)
Returns
true if all parameters were equal, false if not

Definition at line 651 of file SGObject.cpp.

T get ( const Tag< T > &  _tag) const
inherited

Getter for a class parameter, identified by a Tag. Throws an exception if the class does not have such a parameter.

Parameters
_tagname and type information of parameter
Returns
value of the parameter identified by the input tag

Definition at line 367 of file SGObject.h.

T get ( const std::string &  name) const
inherited

Getter for a class parameter, identified by a name. Throws an exception if the class does not have such a parameter.

Parameters
namename of the parameter
Returns
value of the parameter corresponding to the input name and type

Definition at line 388 of file SGObject.h.

SGVector< float64_t > get_alpha ( )
virtualinherited

get alpha vector

Returns
vector to compute posterior mean of Gaussian Process:

\[ \mu = K\alpha \]

where \(\mu\) is the mean and \(K\) is the prior covariance matrix.

Implements CInference.

Definition at line 135 of file SparseInference.cpp.

SGMatrix< float64_t > get_cholesky ( )
virtualinherited

get Cholesky decomposition matrix

Returns
Cholesky decomposition of matrix:

\[ L = Cholesky(sW*K*sW+I) \]

where \(K\) is the prior covariance matrix, \(sW\) is the vector returned by get_diagonal_vector(), and \(I\) is the identity matrix.

Implements CInference.

Definition at line 144 of file SparseInference.cpp.

void * get_derivative_helper ( void *  p)
staticprotectedinherited

pthread helper method to compute negative log marginal likelihood derivatives wrt hyperparameter

Definition at line 268 of file Inference.cpp.

float64_t get_derivative_related_cov ( SGVector< float64_t ddiagKi,
SGMatrix< float64_t dKuui,
SGMatrix< float64_t dKui 
)
protectedvirtual

compute variables which are required to compute negative log marginal likelihood full derivatives wrt cov-like hyperparameter \(\theta\)

Note that scale, which is a hyperparameter in inference_method, is a cov-like hyperparameter hyperparameters in cov function are cov-like hyperparameters

Parameters
ddiagKi\(\textbf{diag}(\frac{\partial {\Sigma_{n}}}{\partial {\theta}})\)
dKuui\(\frac{\partial {\Sigma_{m}}}{\partial {\theta}}\)
dKui\(\frac{\partial {\Sigma_{m,n}}}{\partial {\theta}}\)
Returns
derivative of negative log marginal likelihood

Implements CSingleSparseInference.

Reimplemented in CSingleFITCLaplaceInferenceMethod.

Definition at line 104 of file SingleFITCInference.cpp.

float64_t get_derivative_related_cov ( SGVector< float64_t ddiagKi,
SGMatrix< float64_t dKuui,
SGMatrix< float64_t dKui,
SGVector< float64_t v,
SGMatrix< float64_t R 
)
protectedvirtual

compute variables which are required to compute negative log marginal likelihood derivatives wrt cov-like hyperparameter \(\theta\)

Note that scale, which is a hyperparameter in inference_method, is a cov-like hyperparameter hyperparameters in cov function are cov-like hyperparameters

Parameters
ddiagKi\(\textbf{diag}(\frac{\partial {\Sigma_{n}}}{\partial {\theta}})\)
dKuui\(\frac{\partial {\Sigma_{m}}}{\partial {\theta}}\)
dKui\(\frac{\partial {\Sigma_{m,n}}}{\partial {\theta}}\)
vauxiliary variable related to explicit derivative
Rauxiliary variable related to explicit derivative
Returns
derivative of negative log marginal likelihood

Definition at line 126 of file SingleFITCInference.cpp.

SGVector< float64_t > get_derivative_related_cov_diagonal ( )
protectedvirtual

helper function to compute variables which are required to compute negative log marginal likelihood derivatives wrt the diagonal part of cov-like hyperparameter \(\theta\)

Returns
derivative of negative log marginal likelihood

Definition at line 68 of file SingleFITCInference.cpp.

float64_t get_derivative_related_cov_helper ( SGMatrix< float64_t dKuui,
SGVector< float64_t v,
SGMatrix< float64_t R 
)
protectedvirtual

helper function to compute variables which are required to compute negative log marginal likelihood derivatives wrt cov-like hyperparameter \(\theta\)

Note that scale, which is a hyperparameter in inference_method, is a cov-like hyperparameter hyperparameters in cov function are cov-like hyperparameters what is more, derivative wrt inducing_noise will also use this function

Parameters
dKuui\(\frac{\partial {\Sigma_{m}}}{\partial {\theta}}\)
vauxiliary variable related to explicit derivative
Rauxiliary variable related to explicit derivative
Returns
derivative of negative log marginal likelihood

Definition at line 81 of file SingleFITCInference.cpp.

SGVector< float64_t > get_derivative_related_inducing_features ( SGMatrix< float64_t BdK,
const TParameter param 
)
protectedvirtual

helper function to compute variables which are required to compute negative log marginal likelihood derivatives wrt inducing features

Note that the kernel must support to compute the derivatives wrt inducing features

Parameters
BdKauxiliary variable related to explicit derivative or implicit derivative
paramparameter of given kernel
Returns
derivative of negative log marginal likelihood

Definition at line 210 of file SingleFITCInference.cpp.

float64_t get_derivative_related_mean ( SGVector< float64_t dmu)
protectedvirtual

helper function to compute variables which are required to compute negative log marginal likelihood derivatives wrt mean \(\lambda\)

Parameters
dmu\(\frac{\partial {\mu_{n}}}{\partial {\lambda}}\)
Returns
derivative of negative log marginal likelihood

Reimplemented in CSingleFITCLaplaceInferenceMethod.

Definition at line 149 of file SingleFITCInference.cpp.

SGVector< float64_t > get_derivative_wrt_inducing_features ( const TParameter param)
protectedvirtual

returns derivative of negative log marginal likelihood wrt inducing features (input) Note that in order to call this method, kernel must support FITC inference, which means derivatives wrt inducing features can be computed

Note that the kernel must support to compute the derivatives wrt inducing features

Parameters
paramparameter of given kernel
Returns
derivative of negative log marginal likelihood

Implements CSingleSparseInference.

Reimplemented in CSingleFITCLaplaceInferenceMethod.

Definition at line 258 of file SingleFITCInference.cpp.

SGVector< float64_t > get_derivative_wrt_inducing_noise ( const TParameter param)
protectedvirtual

returns derivative of negative log marginal likelihood wrt inducing noise

Parameters
paramparameter of given inference class
Returns
derivative of negative log marginal likelihood

Implements CSingleSparseInference.

Reimplemented in CSingleFITCLaplaceInferenceMethod.

Definition at line 179 of file SingleFITCInference.cpp.

SGVector< float64_t > get_derivative_wrt_inference_method ( const TParameter param)
protectedvirtualinherited

returns derivative of negative log marginal likelihood wrt parameter of CInference class

Parameters
paramparameter of CInference class
Returns
derivative of negative log marginal likelihood

Implements CSparseInference.

Reimplemented in CSingleFITCLaplaceInferenceMethod.

Definition at line 188 of file SingleSparseInference.cpp.

SGVector< float64_t > get_derivative_wrt_kernel ( const TParameter param)
protectedvirtualinherited

returns derivative of negative log marginal likelihood wrt kernel's parameter

Parameters
paramparameter of given kernel
Returns
derivative of negative log marginal likelihood

Implements CSparseInference.

Reimplemented in CSingleFITCLaplaceInferenceMethod.

Definition at line 240 of file SingleSparseInference.cpp.

virtual SGVector<float64_t> get_derivative_wrt_likelihood_model ( const TParameter param)
protectedpure virtual

returns derivative of negative log marginal likelihood wrt parameter of likelihood model

Parameters
paramparameter of given likelihood model
Returns
derivative of negative log marginal likelihood

Implements CSparseInference.

Implemented in CSingleFITCLaplaceInferenceMethod, and CFITCInferenceMethod.

SGVector< float64_t > get_derivative_wrt_mean ( const TParameter param)
protectedvirtual

returns derivative of negative log marginal likelihood wrt mean function's parameter

Parameters
paramparameter of given mean function
Returns
derivative of negative log marginal likelihood

Implements CSparseInference.

Reimplemented in CSingleFITCLaplaceInferenceMethod.

Definition at line 157 of file SingleFITCInference.cpp.

virtual SGVector<float64_t> get_diagonal_vector ( )
pure virtualinherited
virtual CFeatures* get_features ( )
virtualinherited

get features

Returns
features

Definition at line 266 of file Inference.h.

SGIO * get_global_io ( )
inherited

get the io object

Returns
io object

Definition at line 268 of file SGObject.cpp.

Parallel * get_global_parallel ( )
inherited

get the parallel object

Returns
parallel object

Definition at line 310 of file SGObject.cpp.

Version * get_global_version ( )
inherited

get the version object

Returns
version object

Definition at line 323 of file SGObject.cpp.

virtual CMap<TParameter*, SGVector<float64_t> >* get_gradient ( CMap< TParameter *, CSGObject * > *  parameters)
virtualinherited

get the gradient

Parameters
parametersparameter's dictionary
Returns
map of gradient. Keys are names of parameters, values are values of derivative with respect to that parameter.

Implements CDifferentiableFunction.

Definition at line 245 of file Inference.h.

virtual CFeatures* get_inducing_features ( )
virtualinherited

get inducing features

Returns
features

Definition at line 121 of file SparseInference.h.

float64_t get_inducing_noise ( )
virtualinherited

get the noise for inducing points

Returns
noise noise for inducing points

Definition at line 118 of file SparseInference.cpp.

virtual EInferenceType get_inference_type ( ) const
virtualinherited

return what type of inference we are

Returns
inference type Sparse

Reimplemented from CInference.

Reimplemented in CSingleFITCLaplaceInferenceMethod, CFITCInferenceMethod, and CVarDTCInferenceMethod.

Definition at line 96 of file SparseInference.h.

virtual CKernel* get_kernel ( )
virtualinherited

get kernel

Returns
kernel

Definition at line 283 of file Inference.h.

virtual CLabels* get_labels ( )
virtualinherited

get labels

Returns
labels

Definition at line 317 of file Inference.h.

float64_t get_marginal_likelihood_estimate ( int32_t  num_importance_samples = 1,
float64_t  ridge_size = 1e-15 
)
inherited

Computes an unbiased estimate of the marginal-likelihood (in log-domain),

\[ p(y|X,\theta), \]

where \(y\) are the labels, \(X\) are the features (omitted from in the following expressions), and \(\theta\) represent hyperparameters.

This is done via a Gaussian approximation to the posterior \(q(f|y, \theta)\approx p(f|y, \theta)\), which is computed by the underlying CInference instance (if implemented, otherwise error), and then using an importance sample estimator

\[ p(y|\theta)=\int p(y|f)p(f|\theta)df =\int p(y|f)\frac{p(f|\theta)}{q(f|y, \theta)}q(f|y, \theta)df \approx\frac{1}{n}\sum_{i=1}^n p(y|f^{(i)})\frac{p(f^{(i)}|\theta)} {q(f^{(i)}|y, \theta)}, \]

where \( f^{(i)} \) are samples from the posterior approximation \( q(f|y, \theta) \). The resulting estimator has a low variance if \( q(f|y, \theta) \) is a good approximation. It has large variance otherwise (while still being consistent). Storing all number of log-domain ensures numerical stability.

Parameters
num_importance_samplesthe number of importance samples \(n\) from \( q(f|y, \theta) \).
ridge_sizescalar that is added to the diagonal of the involved Gaussian distribution's covariance of GP prior and posterior approximation to stabilise things. Increase if covariance matrix is not numerically positive semi-definite.
Returns
unbiased estimate of the marginal likelihood function \( p(y|\theta),\) in log-domain.

Definition at line 139 of file Inference.cpp.

virtual CMeanFunction* get_mean ( )
virtualinherited

get mean

Returns
mean

Definition at line 300 of file Inference.h.

CLikelihoodModel* get_model ( )
inherited

get likelihood model

Returns
likelihood

Definition at line 334 of file Inference.h.

SGStringList< char > get_modelsel_names ( )
inherited
Returns
vector of names of all parameters which are registered for model selection

Definition at line 531 of file SGObject.cpp.

char * get_modsel_param_descr ( const char *  param_name)
inherited

Returns description of a given parameter string, if it exists. SG_ERROR otherwise

Parameters
param_namename of the parameter
Returns
description of the parameter

Definition at line 555 of file SGObject.cpp.

index_t get_modsel_param_index ( const char *  param_name)
inherited

Returns index of model selection parameter with provided index

Parameters
param_namename of model selection parameter
Returns
index of model selection parameter with provided name, -1 if there is no such

Definition at line 568 of file SGObject.cpp.

SGMatrix< float64_t > get_multiclass_E ( )
virtualinherited

get the E matrix used for multi classification

Returns
the matrix for multi classification

Definition at line 71 of file Inference.cpp.

virtual const char* get_name ( ) const
virtual

returns the name of the inference method

Returns
name SingleFITCInference

Reimplemented from CSingleSparseInference.

Reimplemented in CSingleFITCLaplaceInferenceMethod, and CFITCInferenceMethod.

Definition at line 93 of file SingleFITCInference.h.

virtual float64_t get_negative_log_marginal_likelihood ( )
pure virtualinherited

get negative log marginal likelihood

Returns
the negative log of the marginal likelihood function:

\[ -log(p(y|X, \theta)) \]

where \(y\) are the labels, \(X\) are the features, and \(\theta\) represent hyperparameters.

Implemented in CSingleFITCLaplaceInferenceMethod, CMultiLaplaceInferenceMethod, CKLInference, CExactInferenceMethod, CFITCInferenceMethod, CVarDTCInferenceMethod, CEPInferenceMethod, and CSingleLaplaceInferenceMethod.

CMap< TParameter *, SGVector< float64_t > > * get_negative_log_marginal_likelihood_derivatives ( CMap< TParameter *, CSGObject * > *  parameters)
virtualinherited

get log marginal likelihood gradient

Returns
vector of the marginal likelihood function gradient with respect to hyperparameters (under the current approximation to the posterior \(q(f|y)\approx p(f|y)\):

\[ -\frac{\partial log(p(y|X, \theta))}{\partial \theta} \]

where \(y\) are the labels, \(X\) are the features, and \(\theta\) represent hyperparameters.

Definition at line 198 of file Inference.cpp.

virtual SGMatrix<float64_t> get_posterior_covariance ( )
pure virtualinherited

returns covariance matrix \(\Sigma\) of the Gaussian distribution \(\mathcal{N}(\mu,\Sigma)\), which is an approximation to the posterior:

\[ p(f|y) \approx q(f|y) = \mathcal{N}(\mu,\Sigma) \]

in case if particular inference method doesn't compute posterior \(p(f|y)\) exactly, and it returns covariance matrix \(\Sigma\) of the posterior Gaussian distribution \(\mathcal{N}(\mu,\Sigma)\) otherwise.

Returns
covariance matrix

Implements CInference.

Implemented in CFITCInferenceMethod, CVarDTCInferenceMethod, and CSingleFITCLaplaceInferenceMethod.

virtual SGVector<float64_t> get_posterior_mean ( )
pure virtualinherited

returns mean vector \(\mu\) of the Gaussian distribution \(\mathcal{N}(\mu,\Sigma)\), which is an approximation to the posterior:

\[ p(f|y) \approx q(f|y) = \mathcal{N}(\mu,\Sigma) \]

in case if particular inference method doesn't compute posterior \(p(f|y)\) exactly, and it returns covariance matrix \(\Sigma\) of the posterior Gaussian distribution \(\mathcal{N}(\mu,\Sigma)\) otherwise.

Returns
mean vector

Implements CInference.

Implemented in CFITCInferenceMethod, CVarDTCInferenceMethod, and CSingleFITCLaplaceInferenceMethod.

float64_t get_scale ( ) const
virtualinherited

get kernel scale

Returns
kernel scale

Definition at line 60 of file Inference.cpp.

virtual SGVector<float64_t> get_value ( )
virtualinherited

get the function value

Returns
vector that represents the function value

Implements CDifferentiableFunction.

Definition at line 255 of file Inference.h.

bool has ( const std::string &  name) const
inherited

Checks if object has a class parameter identified by a name.

Parameters
namename of the parameter
Returns
true if the parameter exists with the input name

Definition at line 289 of file SGObject.h.

bool has ( const Tag< T > &  tag) const
inherited

Checks if object has a class parameter identified by a Tag.

Parameters
tagtag of the parameter containing name and type information
Returns
true if the parameter exists with the input tag

Definition at line 301 of file SGObject.h.

bool has ( const std::string &  name) const
inherited

Checks if a type exists for a class parameter identified by a name.

Parameters
namename of the parameter
Returns
true if the parameter exists with the input name and type

Definition at line 312 of file SGObject.h.

bool is_generic ( EPrimitiveType *  generic) const
virtualinherited

If the SGSerializable is a class template then TRUE will be returned and GENERIC is set to the type of the generic.

Parameters
genericset to the type of the generic if returning TRUE
Returns
TRUE if a class template.

Definition at line 329 of file SGObject.cpp.

bool load_serializable ( CSerializableFile file,
const char *  prefix = "" 
)
virtualinherited

Load this object from file. If it will fail (returning FALSE) then this object will contain inconsistent data and should not be used!

Parameters
filewhere to load from
prefixprefix for members
Returns
TRUE if done, otherwise FALSE

Definition at line 402 of file SGObject.cpp.

void load_serializable_post ( )
throw (ShogunException
)
protectedvirtualinherited

Can (optionally) be overridden to post-initialize some member variables which are not PARAMETER::ADD'ed. Make sure that at first the overridden method BASE_CLASS::LOAD_SERIALIZABLE_POST is called.

Exceptions
ShogunExceptionwill be thrown if an error occurs.

Reimplemented in CKernel, CWeightedDegreePositionStringKernel, CList, CAlphabet, CLinearHMM, CGaussianKernel, CInverseMultiQuadricKernel, CCircularKernel, and CExponentialKernel.

Definition at line 459 of file SGObject.cpp.

void load_serializable_pre ( )
throw (ShogunException
)
protectedvirtualinherited

Can (optionally) be overridden to pre-initialize some member variables which are not PARAMETER::ADD'ed. Make sure that at first the overridden method BASE_CLASS::LOAD_SERIALIZABLE_PRE is called.

Exceptions
ShogunExceptionwill be thrown if an error occurs.

Reimplemented in CDynamicArray< T >, CDynamicArray< float64_t >, CDynamicArray< float32_t >, CDynamicArray< int32_t >, CDynamicArray< char >, CDynamicArray< bool >, and CDynamicObjectArray.

Definition at line 454 of file SGObject.cpp.

void optimize_inducing_features ( )
virtualinherited

opitmize inducing features

Definition at line 357 of file SingleSparseInference.cpp.

bool parameter_hash_changed ( )
virtualinherited
Returns
whether parameter combination has changed since last update

Definition at line 295 of file SGObject.cpp.

void print_modsel_params ( )
inherited

prints all parameter registered for model selection and their type

Definition at line 507 of file SGObject.cpp.

void print_serializable ( const char *  prefix = "")
virtualinherited

prints registered parameters out

Parameters
prefixprefix for members

Definition at line 341 of file SGObject.cpp.

void register_minimizer ( Minimizer minimizer)
virtualinherited

Set a minimizer

Parameters
minimizerminimizer used in inference method

Reimplemented in CEPInferenceMethod, CKLInference, CExactInferenceMethod, CKLDualInferenceMethod, CSingleFITCLaplaceInferenceMethod, CFITCInferenceMethod, CVarDTCInferenceMethod, and CSingleLaplaceInferenceMethod.

Definition at line 128 of file Inference.cpp.

void register_param ( Tag< T > &  _tag,
const T &  value 
)
protectedinherited

Registers a class parameter which is identified by a tag. This enables the parameter to be modified by set() and retrieved by get(). Parameters can be registered in the constructor of the class.

Parameters
_tagname and type information of parameter
valuevalue of the parameter

Definition at line 439 of file SGObject.h.

void register_param ( const std::string &  name,
const T &  value 
)
protectedinherited

Registers a class parameter which is identified by a name. This enables the parameter to be modified by set() and retrieved by get(). Parameters can be registered in the constructor of the class.

Parameters
namename of the parameter
valuevalue of the parameter along with type information

Definition at line 452 of file SGObject.h.

bool save_serializable ( CSerializableFile file,
const char *  prefix = "" 
)
virtualinherited

Save this object to file.

Parameters
filewhere to save the object; will be closed during returning if PREFIX is an empty string.
prefixprefix for members
Returns
TRUE if done, otherwise FALSE

Definition at line 347 of file SGObject.cpp.

void save_serializable_post ( )
throw (ShogunException
)
protectedvirtualinherited

Can (optionally) be overridden to post-initialize some member variables which are not PARAMETER::ADD'ed. Make sure that at first the overridden method BASE_CLASS::SAVE_SERIALIZABLE_POST is called.

Exceptions
ShogunExceptionwill be thrown if an error occurs.

Reimplemented in CKernel.

Definition at line 469 of file SGObject.cpp.

void save_serializable_pre ( )
throw (ShogunException
)
protectedvirtualinherited

Can (optionally) be overridden to pre-initialize some member variables which are not PARAMETER::ADD'ed. Make sure that at first the overridden method BASE_CLASS::SAVE_SERIALIZABLE_PRE is called.

Exceptions
ShogunExceptionwill be thrown if an error occurs.

Reimplemented in CKernel, CDynamicArray< T >, CDynamicArray< float64_t >, CDynamicArray< float32_t >, CDynamicArray< int32_t >, CDynamicArray< char >, CDynamicArray< bool >, and CDynamicObjectArray.

Definition at line 464 of file SGObject.cpp.

void set ( const Tag< T > &  _tag,
const T &  value 
)
inherited

Setter for a class parameter, identified by a Tag. Throws an exception if the class does not have such a parameter.

Parameters
_tagname and type information of parameter
valuevalue of the parameter

Definition at line 328 of file SGObject.h.

void set ( const std::string &  name,
const T &  value 
)
inherited

Setter for a class parameter, identified by a name. Throws an exception if the class does not have such a parameter.

Parameters
namename of the parameter
valuevalue of the parameter along with type information

Definition at line 354 of file SGObject.h.

virtual void set_features ( CFeatures feat)
virtualinherited

set features

Parameters
featfeatures to set

Definition at line 272 of file Inference.h.

void set_generic ( )
inherited

Definition at line 74 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 79 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 84 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 89 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 94 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 99 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 104 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 109 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 114 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 119 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 124 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 129 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 134 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 139 of file SGObject.cpp.

void set_generic ( )
inherited

Definition at line 144 of file SGObject.cpp.

void set_generic ( )
inherited

set generic type to T

void set_global_io ( SGIO io)
inherited

set the io object

Parameters
ioio object to use

Definition at line 261 of file SGObject.cpp.

void set_global_parallel ( Parallel parallel)
inherited

set the parallel object

Parameters
parallelparallel object to use

Definition at line 274 of file SGObject.cpp.

void set_global_version ( Version version)
inherited

set the version object

Parameters
versionversion object to use

Definition at line 316 of file SGObject.cpp.

virtual void set_inducing_features ( CFeatures feat)
virtualinherited

set inducing features

Parameters
featfeatures to set

Definition at line 108 of file SparseInference.h.

void set_inducing_noise ( float64_t  noise)
virtualinherited

set the noise for inducing points

Parameters
noisenoise for inducing points

The noise is used to enfore the kernel matrix about the inducing points are positive definite

Definition at line 112 of file SparseInference.cpp.

void set_kernel ( CKernel kern)
virtualinherited

set kernel

Parameters
kernkernel to set

Reimplemented from CInference.

Definition at line 164 of file SingleSparseInference.cpp.

virtual void set_labels ( CLabels lab)
virtualinherited

set labels

Parameters
lablabel to set

Definition at line 323 of file Inference.h.

void set_lower_bound_of_inducing_features ( SGVector< float64_t bound)
virtualinherited

set the lower bound of inducing features

Parameters
boundlower bound constrains of inducing features

Note that if the length of the bound is 1, it means the bound constraint applies to each dimension of all inducing features

Note that if the length of the bound is greater than 1, it means each dimension of the bound constraint applies to the corresponding dimension of inducing features

Definition at line 299 of file SingleSparseInference.cpp.

void set_max_iterations_for_inducing_features ( int32_t  it)
virtualinherited

set the max number of iterations used in optimization of inducing features

Parameters
itmax number of iterations

Definition at line 310 of file SingleSparseInference.cpp.

virtual void set_mean ( CMeanFunction m)
virtualinherited

set mean

Parameters
mmean function to set

Definition at line 306 of file Inference.h.

virtual void set_model ( CLikelihoodModel mod)
virtualinherited

set likelihood model

Parameters
modmodel to set

Reimplemented in CKLDualInferenceMethod, and CKLInference.

Definition at line 340 of file Inference.h.

void set_scale ( float64_t  scale)
virtualinherited

set kernel scale

Parameters
scalescale to be set

Definition at line 65 of file Inference.cpp.

void set_tolearance_for_inducing_features ( float64_t  tol)
virtualinherited

set the tolearance used in optimization of inducing features

Parameters
toltolearance

Definition at line 315 of file SingleSparseInference.cpp.

void set_upper_bound_of_inducing_features ( SGVector< float64_t bound)
virtualinherited

set the upper bound of inducing features

Parameters
boundupper bound constrains of inducing features

Note that if the length of the bound is 1, it means the bound constraint applies to each dimension of all inducing features

Note that if the length of the bound is greater than 1, it means each dimension of the bound constraint applies to the corresponding dimension of inducing features

Definition at line 304 of file SingleSparseInference.cpp.

CSGObject * shallow_copy ( ) const
virtualinherited

A shallow copy. All the SGObject instance variables will be simply assigned and SG_REF-ed.

Reimplemented in CGaussianKernel.

Definition at line 225 of file SGObject.cpp.

virtual bool supports_binary ( ) const
virtualinherited

whether combination of inference method and given likelihood function supports binary classification

Returns
false

Reimplemented in CEPInferenceMethod, CKLInference, CSingleFITCLaplaceInferenceMethod, and CSingleLaplaceInferenceMethod.

Definition at line 371 of file Inference.h.

virtual bool supports_multiclass ( ) const
virtualinherited

whether combination of inference method and given likelihood function supports multiclass classification

Returns
false

Reimplemented in CMultiLaplaceInferenceMethod.

Definition at line 378 of file Inference.h.

virtual bool supports_regression ( ) const
virtualinherited

whether combination of inference method and given likelihood function supports regression

Returns
false

Reimplemented in CExactInferenceMethod, CKLInference, CFITCInferenceMethod, CVarDTCInferenceMethod, CSingleFITCLaplaceInferenceMethod, and CSingleLaplaceInferenceMethod.

Definition at line 364 of file Inference.h.

void unset_generic ( )
inherited

unset generic type

this has to be called in classes specializing a template class

Definition at line 336 of file SGObject.cpp.

virtual void update ( )
pure virtualinherited

update all matrices

Reimplemented from CInference.

Implemented in CFITCInferenceMethod, CVarDTCInferenceMethod, and CSingleFITCLaplaceInferenceMethod.

virtual void update_alpha ( )
protectedpure virtual

update alpha vector

Implements CInference.

Implemented in CSingleFITCLaplaceInferenceMethod, and CFITCInferenceMethod.

virtual void update_chol ( )
protectedpure virtual

update cholesky matrix

Implements CInference.

Implemented in CSingleFITCLaplaceInferenceMethod, and CFITCInferenceMethod.

virtual void update_deriv ( )
protectedpure virtual

update matrices which are required to compute negative log marginal likelihood derivatives wrt hyperparameter

Implements CInference.

Implemented in CSingleFITCLaplaceInferenceMethod, and CFITCInferenceMethod.

void update_parameter_hash ( )
virtualinherited

Updates the hash of current parameter combination

Definition at line 281 of file SGObject.cpp.

void update_train_kernel ( )
protectedvirtualinherited

update train kernel matrix

Reimplemented from CInference.

Definition at line 153 of file SparseInference.cpp.

Member Data Documentation

SGIO* io
inherited

io

Definition at line 537 of file SGObject.h.

SGVector<float64_t> m_al
protected

Note that alpha is NOT post.alpha alpha and post.alpha are defined in infFITC.m and infFITC_Laplace.m

Definition at line 230 of file SingleFITCInference.h.

SGVector<float64_t> m_alpha
protectedinherited

alpha vector used in process mean calculation

Definition at line 484 of file Inference.h.

SGMatrix<float64_t> m_B
protected

B is defined in infFITC.m and infFITC_Laplace.m

Definition at line 239 of file SingleFITCInference.h.

SGMatrix<float64_t> m_E
protectedinherited

the matrix used for multi classification

Definition at line 496 of file Inference.h.

CFeatures* m_features
protectedinherited

features to use

Definition at line 478 of file Inference.h.

bool m_fully_sparse
protectedinherited

whether the kernel supports to get the gradient wrt inducing points or not

Definition at line 224 of file SingleSparseInference.h.

Parameter* m_gradient_parameters
inherited

parameters wrt which we can compute gradients

Definition at line 552 of file SGObject.h.

bool m_gradient_update
protectedinherited

Whether gradients are updated

Definition at line 499 of file Inference.h.

uint32_t m_hash
inherited

Hash of parameter values

Definition at line 555 of file SGObject.h.

float64_t m_ind_tolerance
protectedinherited

tolearance used in optimizing inducing_features

Definition at line 197 of file SingleSparseInference.h.

SGMatrix<float64_t> m_inducing_features
protectedinherited

inducing features for approximation

Definition at line 304 of file SparseInference.h.

FirstOrderMinimizer* m_inducing_minimizer
protectedinherited

minimizer used in finding optimal inducing features

Definition at line 230 of file SingleSparseInference.h.

CKernel* m_kernel
protectedinherited

covariance function

Definition at line 469 of file Inference.h.

SGMatrix<float64_t> m_ktrtr
protectedinherited

kernel matrix from features (non-scalled by inference scalling)

Definition at line 493 of file Inference.h.

SGVector<float64_t> m_ktrtr_diag
protectedinherited

diagonal elements of kernel matrix m_ktrtr

Definition at line 322 of file SparseInference.h.

SGMatrix<float64_t> m_ktru
protectedinherited

covariance matrix of inducing features and training features

Definition at line 313 of file SparseInference.h.

SGMatrix<float64_t> m_kuu
protectedinherited

covariance matrix of inducing features

Definition at line 310 of file SparseInference.h.

SGMatrix<float64_t> m_L
protectedinherited

upper triangular factor of Cholesky decomposition

Definition at line 487 of file Inference.h.

CLabels* m_labels
protectedinherited

labels of features

Definition at line 481 of file Inference.h.

CLock* m_lock
protectedinherited

a lock used to parallelly compute derivatives wrt hyperparameters

Definition at line 227 of file SingleSparseInference.h.

float64_t m_log_ind_noise
protectedinherited

noise of the inducing variables

Definition at line 307 of file SparseInference.h.

float64_t m_log_scale
protectedinherited

kernel scale

Definition at line 490 of file Inference.h.

SGVector<float64_t> m_lower_bound
protectedinherited

lower bound of inducing features

Definition at line 188 of file SingleSparseInference.h.

float64_t m_max_ind_iterations
protectedinherited

max number of iterations

Definition at line 194 of file SingleSparseInference.h.

CMeanFunction* m_mean
protectedinherited

mean function

Definition at line 472 of file Inference.h.

Minimizer* m_minimizer
protectedinherited

minimizer

Definition at line 466 of file Inference.h.

CLikelihoodModel* m_model
protectedinherited

likelihood function to use

Definition at line 475 of file Inference.h.

Parameter* m_model_selection_parameters
inherited

model selection parameters

Definition at line 549 of file SGObject.h.

SGVector<float64_t> m_mu
protectedinherited

mean vector of the the posterior Gaussian distribution

Definition at line 319 of file SparseInference.h.

bool m_opt_inducing_features
protectedinherited

whether optimize inducing features

Definition at line 200 of file SingleSparseInference.h.

Parameter* m_parameters
inherited

parameters

Definition at line 546 of file SGObject.h.

SGMatrix<float64_t> m_Rvdd
protected

Rvdd=W where W is defined in infFITC.m and Rvdd is defined in infFITC_Laplace.m Note that W is NOT the diagonal matrix

Definition at line 248 of file SingleFITCInference.h.

SGMatrix<float64_t> m_Sigma
protectedinherited

covariance matrix of the the posterior Gaussian distribution

Definition at line 316 of file SparseInference.h.

SGVector<float64_t> m_t
protected

t=1/g_sn2 in regression, where g_sn2 is defined in infFITC.m t=W.*dd in Laplace for binary classification, where W and dd are defined in infFITC_Laplace.m

Definition at line 236 of file SingleFITCInference.h.

SGVector<float64_t> m_upper_bound
protectedinherited

upper bound of inducing features

Definition at line 191 of file SingleSparseInference.h.

SGMatrix<float64_t> m_V
protected

V defined in infFITC.m and infFITC_Laplace.m

Definition at line 251 of file SingleFITCInference.h.

SGVector<float64_t> m_w
protected

w=B*al

Definition at line 242 of file SingleFITCInference.h.

Parallel* parallel
inherited

parallel

Definition at line 540 of file SGObject.h.

Version* version
inherited

version

Definition at line 543 of file SGObject.h.


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

SHOGUN Machine Learning Toolbox - Documentation