SHOGUN  v3.2.0
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines
Public Member Functions | Static Public Member Functions | Public Attributes | Protected Member Functions | Protected Attributes | Friends
CMKL Class Reference

Detailed Description

Multiple Kernel Learning.

A support vector machine based method for use with multiple kernels. In Multiple Kernel Learning (MKL) in addition to the SVM \(\bf\alpha\) and bias term \(b\) the kernel weights \(\bf\beta\) are estimated in training. The resulting kernel method can be stated as

\[ f({\bf x})=\sum_{i=0}^{N-1} \alpha_i \sum_{j=0}^M \beta_j k_j({\bf x}, {\bf x_i})+b . \]

where \(N\) is the number of training examples \(\alpha_i\) are the weights assigned to each training example \(\beta_j\) are the weights assigned to each sub-kernel \(k_j(x,x')\) are sub-kernels and \(b\) the bias.

Kernels have to be chosen a-priori. In MKL \(\alpha_i,\;\beta\) and bias are determined by solving the following optimization program

\begin{eqnarray*} \mbox{min} && \gamma-\sum_{i=1}^N\alpha_i\\ \mbox{w.r.t.} && \gamma\in R, \alpha\in R^N \nonumber\\ \mbox{s.t.} && {\bf 0}\leq\alpha\leq{\bf 1}C,\;\;\sum_{i=1}^N \alpha_i y_i=0 \nonumber\\ && \frac{1}{2}\sum_{i,j=1}^N \alpha_i \alpha_j y_i y_j k_k({\bf x}_i,{\bf x}_j)\leq \gamma,\;\; \forall k=1,\ldots,K\nonumber\\ \end{eqnarray*}

here C is a pre-specified regularization parameter.

Within shogun this optimization problem is solved using semi-infinite programming. For 1-norm MKL using one of the two approaches described in

Soeren Sonnenburg, Gunnar Raetsch, Christin Schaefer, and Bernhard Schoelkopf. Large Scale Multiple Kernel Learning. Journal of Machine Learning Research, 7:1531-1565, July 2006.

The first approach (also called the wrapper algorithm) wraps around a single kernel SVMs, alternatingly solving for \(\alpha\) and \(\beta\). It is using a traditional SVM to generate new violated constraints and thus requires a single kernel SVM and any of the SVMs contained in shogun can be used. In the MKL step either a linear program is solved via glpk or cplex or analytically or a newton (for norms>1) step is performed.

The second much faster but also more memory demanding approach performing interleaved optimization, is integrated into the chunking-based SVMlight.

In addition sparsity of MKL can be controlled by the choice of the \(L_p\)-norm regularizing \(\beta\) as described in

Marius Kloft, Ulf Brefeld, Soeren Sonnenburg, and Alexander Zien. Efficient and accurate lp-norm multiple kernel learning. In Advances in Neural Information Processing Systems 21. MIT Press, Cambridge, MA, 2009.

An alternative way to control the sparsity is the elastic-net regularization, which can be formulated into the following optimization problem:

\begin{eqnarray*} \mbox{min} && C\sum_{i=1}^N\ell\left(\sum_{k=1}^Kf_k(x_i)+b,y_i\right)+(1-\lambda)\left(\sum_{k=1}^K\|f_k\|_{\mathcal{H}_k}\right)^2+\lambda\sum_{k=1}^K\|f_k\|_{\mathcal{H}_k}^2\\ \mbox{w.r.t.} && f_1\in\mathcal{H}_1,f_2\in\mathcal{H}_2,\ldots,f_K\in\mathcal{H}_K,\,b\in R \nonumber\\ \end{eqnarray*}

where \(\ell\) is a loss function. Here \(\lambda\) controls the trade-off between the two regularization terms. \(\lambda=0\) corresponds to \(L_1\)-MKL, whereas \(\lambda=1\) corresponds to the uniform-weighted combination of kernels ( \(L_\infty\)-MKL). This approach was studied by Shawe-Taylor (2008) "Kernel Learning for Novelty Detection" (NIPS MKL Workshop 2008) and Tomioka & Suzuki (2009) "Sparsity-accuracy trade-off in MKL" (NIPS MKL Workshop 2009).

Definition at line 95 of file MKL.h.

Inheritance diagram for CMKL:
Inheritance graph
[legend]

List of all members.

Public Member Functions

 CMKL (CSVM *s=NULL)
virtual ~CMKL ()
void set_constraint_generator (CSVM *s)
void set_svm (CSVM *s)
CSVMget_svm ()
void set_C_mkl (float64_t C)
void set_mkl_norm (float64_t norm)
void set_elasticnet_lambda (float64_t elasticnet_lambda)
void set_mkl_block_norm (float64_t q)
void set_interleaved_optimization_enabled (bool enable)
bool get_interleaved_optimization_enabled ()
float64_t compute_mkl_primal_objective ()
virtual float64_t compute_mkl_dual_objective ()
float64_t compute_elasticnet_dual_objective ()
void set_mkl_epsilon (float64_t eps)
float64_t get_mkl_epsilon ()
int32_t get_mkl_iterations ()
virtual bool perform_mkl_step (const float64_t *sumw, float64_t suma)
virtual float64_t compute_sum_alpha ()=0
virtual void compute_sum_beta (float64_t *sumw)
virtual const char * get_name () const
 MACHINE_PROBLEM_TYPE (PT_BINARY)
void set_defaults (int32_t num_sv=0)
virtual SGVector< float64_tget_linear_term ()
virtual void set_linear_term (const SGVector< float64_t > linear_term)
bool load (FILE *svm_file)
bool save (FILE *svm_file)
void set_nu (float64_t nue)
void set_C (float64_t c_neg, float64_t c_pos)
void set_epsilon (float64_t eps)
void set_tube_epsilon (float64_t eps)
float64_t get_tube_epsilon ()
void set_qpsize (int32_t qps)
float64_t get_epsilon ()
float64_t get_nu ()
float64_t get_C1 ()
float64_t get_C2 ()
int32_t get_qpsize ()
void set_shrinking_enabled (bool enable)
bool get_shrinking_enabled ()
float64_t compute_svm_dual_objective ()
float64_t compute_svm_primal_objective ()
void set_objective (float64_t v)
float64_t get_objective ()
void set_callback_function (CMKL *m, bool(*cb)(CMKL *mkl, const float64_t *sumw, const float64_t suma))
void set_kernel (CKernel *k)
CKernelget_kernel ()
void set_batch_computation_enabled (bool enable)
bool get_batch_computation_enabled ()
void set_linadd_enabled (bool enable)
bool get_linadd_enabled ()
void set_bias_enabled (bool enable_bias)
bool get_bias_enabled ()
float64_t get_bias ()
void set_bias (float64_t bias)
int32_t get_support_vector (int32_t idx)
float64_t get_alpha (int32_t idx)
bool set_support_vector (int32_t idx, int32_t val)
bool set_alpha (int32_t idx, float64_t val)
int32_t get_num_support_vectors ()
void set_alphas (SGVector< float64_t > alphas)
void set_support_vectors (SGVector< int32_t > svs)
SGVector< int32_t > get_support_vectors ()
SGVector< float64_tget_alphas ()
bool create_new_model (int32_t num)
bool init_kernel_optimization ()
virtual CRegressionLabelsapply_regression (CFeatures *data=NULL)
virtual CBinaryLabelsapply_binary (CFeatures *data=NULL)
virtual float64_t apply_one (int32_t num)
virtual bool train_locked (SGVector< index_t > indices)
virtual CBinaryLabelsapply_locked_binary (SGVector< index_t > indices)
virtual CRegressionLabelsapply_locked_regression (SGVector< index_t > indices)
virtual SGVector< float64_tapply_locked_get_output (SGVector< index_t > indices)
virtual void data_lock (CLabels *labs, CFeatures *features=NULL)
virtual void data_unlock ()
virtual bool supports_locking () const
virtual bool train (CFeatures *data=NULL)
virtual CLabelsapply (CFeatures *data=NULL)
virtual CMulticlassLabelsapply_multiclass (CFeatures *data=NULL)
virtual CStructuredLabelsapply_structured (CFeatures *data=NULL)
virtual CLatentLabelsapply_latent (CFeatures *data=NULL)
virtual void set_labels (CLabels *lab)
virtual CLabelsget_labels ()
void set_max_train_time (float64_t t)
float64_t get_max_train_time ()
virtual EMachineType get_classifier_type ()
void set_solver_type (ESolverType st)
ESolverType get_solver_type ()
virtual void set_store_model_features (bool store_model)
virtual CLabelsapply_locked (SGVector< index_t > indices)
virtual CMulticlassLabelsapply_locked_multiclass (SGVector< index_t > indices)
virtual CStructuredLabelsapply_locked_structured (SGVector< index_t > indices)
virtual CLatentLabelsapply_locked_latent (SGVector< index_t > indices)
virtual void post_lock (CLabels *labs, CFeatures *features)
bool is_data_locked () const
virtual EProblemType get_machine_problem_type () const
virtual CSGObjectshallow_copy () const
virtual CSGObjectdeep_copy () const
virtual bool is_generic (EPrimitiveType *generic) const
template<class T >
void set_generic ()
void unset_generic ()
virtual void print_serializable (const char *prefix="")
virtual bool save_serializable (CSerializableFile *file, const char *prefix="", int32_t param_version=Version::get_version_parameter())
virtual bool load_serializable (CSerializableFile *file, const char *prefix="", int32_t param_version=Version::get_version_parameter())
DynArray< TParameter * > * load_file_parameters (const SGParamInfo *param_info, int32_t file_version, CSerializableFile *file, const char *prefix="")
DynArray< TParameter * > * load_all_file_parameters (int32_t file_version, int32_t current_version, CSerializableFile *file, const char *prefix="")
void map_parameters (DynArray< TParameter * > *param_base, int32_t &base_version, DynArray< const SGParamInfo * > *target_param_infos)
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)
virtual bool update_parameter_hash ()
virtual bool equals (CSGObject *other, float64_t accuracy=0.0)
virtual CSGObjectclone ()

Static Public Member Functions

static bool perform_mkl_step_helper (CMKL *mkl, const float64_t *sumw, const float64_t suma)
static void * apply_helper (void *p)

Public Attributes

SGIOio
Parallelparallel
Versionversion
Parameterm_parameters
Parameterm_model_selection_parameters
Parameterm_gradient_parameters
ParameterMapm_parameter_map
uint32_t m_hash

Protected Member Functions

virtual bool train_machine (CFeatures *data=NULL)
virtual void init_training ()=0
void perform_mkl_step (float64_t *beta, float64_t *old_beta, int num_kernels, int32_t *label, int32_t *active2dnum, float64_t *a, float64_t *lin, float64_t *sumw, int32_t &inner_iters)
float64_t compute_optimal_betas_via_cplex (float64_t *beta, const float64_t *old_beta, int32_t num_kernels, const float64_t *sumw, float64_t suma, int32_t &inner_iters)
float64_t compute_optimal_betas_via_glpk (float64_t *beta, const float64_t *old_beta, int num_kernels, const float64_t *sumw, float64_t suma, int32_t &inner_iters)
float64_t compute_optimal_betas_elasticnet (float64_t *beta, const float64_t *old_beta, const int32_t num_kernels, const float64_t *sumw, const float64_t suma, const float64_t mkl_objective)
void elasticnet_transform (float64_t *beta, float64_t lmd, int32_t len)
void elasticnet_dual (float64_t *ff, float64_t *gg, float64_t *hh, const float64_t &del, const float64_t *nm, int32_t len, const float64_t &lambda)
float64_t compute_optimal_betas_directly (float64_t *beta, const float64_t *old_beta, const int32_t num_kernels, const float64_t *sumw, const float64_t suma, const float64_t mkl_objective)
float64_t compute_optimal_betas_block_norm (float64_t *beta, const float64_t *old_beta, const int32_t num_kernels, const float64_t *sumw, const float64_t suma, const float64_t mkl_objective)
float64_t compute_optimal_betas_newton (float64_t *beta, const float64_t *old_beta, int32_t num_kernels, const float64_t *sumw, float64_t suma, float64_t mkl_objective)
virtual bool converged ()
void init_solver ()
bool init_cplex ()
void set_qnorm_constraints (float64_t *beta, int32_t num_kernels)
bool cleanup_cplex ()
bool init_glpk ()
bool cleanup_glpk ()
bool check_glp_status (glp_prob *lp)
virtual float64_tget_linear_term_array ()
SGVector< float64_tapply_get_outputs (CFeatures *data)
virtual void store_model_features ()
virtual bool is_label_valid (CLabels *lab) const
virtual bool train_require_labels () const
virtual TParametermigrate (DynArray< TParameter * > *param_base, const SGParamInfo *target)
virtual void one_to_one_migration_prepare (DynArray< TParameter * > *param_base, const SGParamInfo *target, TParameter *&replacement, TParameter *&to_migrate, char *old_name=NULL)
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)

Protected Attributes

CSVMsvm
float64_t C_mkl
float64_t mkl_norm
float64_t ent_lambda
float64_t mkl_block_norm
float64_tbeta_local
int32_t mkl_iterations
float64_t mkl_epsilon
bool interleaved_optimization
float64_tW
float64_t w_gap
float64_t rho
CTime training_time_clock
CPXENVptr env
CPXLPptr lp_cplex
glp_prob * lp_glpk
glp_smcp * lp_glpk_parm
bool lp_initialized
SGVector< float64_tm_linear_term
bool svm_loaded
float64_t epsilon
float64_t tube_epsilon
float64_t nu
float64_t C1
float64_t C2
float64_t objective
int32_t qpsize
bool use_shrinking
bool(* callback )(CMKL *mkl, const float64_t *sumw, const float64_t suma)
CMKLmkl
CKernelkernel
CCustomKernelm_custom_kernel
CKernelm_kernel_backup
bool use_batch_computation
bool use_linadd
bool use_bias
float64_t m_bias
SGVector< float64_tm_alpha
SGVector< int32_t > m_svs
float64_t m_max_train_time
CLabelsm_labels
ESolverType m_solver_type
bool m_store_model_features
bool m_data_locked

Friends

class CMulticlassSVM

Constructor & Destructor Documentation

CMKL ( CSVM s = NULL)

Constructor

Parameters:
sSVM to use as constraint generator in MKL SIP

Definition at line 21 of file MKL.cpp.

~CMKL ( ) [virtual]

Destructor

Definition at line 40 of file MKL.cpp.


Member Function Documentation

CLabels * apply ( CFeatures data = NULL) [virtual, inherited]

apply machine to data if data is not specified apply to the current features

Parameters:
data(test)data to be classified
Returns:
classified labels

Definition at line 162 of file Machine.cpp.

CBinaryLabels * apply_binary ( CFeatures data = NULL) [virtual, inherited]

apply kernel machine to data for binary classification task

Parameters:
data(test)data to be classified
Returns:
classified labels

Reimplemented from CMachine.

Reimplemented in CDomainAdaptationSVM.

Definition at line 245 of file KernelMachine.cpp.

SGVector< float64_t > apply_get_outputs ( CFeatures data) [protected, inherited]

apply get outputs

Parameters:
datafeatures to compute outputs
Returns:
outputs

Definition at line 251 of file KernelMachine.cpp.

void * apply_helper ( void *  p) [static, inherited]

apply example helper, used in threads

Parameters:
pparams of the thread
Returns:
nothing really

Definition at line 421 of file KernelMachine.cpp.

CLatentLabels * apply_latent ( CFeatures data = NULL) [virtual, inherited]

apply machine to data in means of latent problem

Reimplemented in CLinearLatentMachine.

Definition at line 242 of file Machine.cpp.

CLabels * apply_locked ( SGVector< index_t indices) [virtual, inherited]

Applies a locked machine on a set of indices. Error if machine is not locked

Parameters:
indicesindex vector (of locked features) that is predicted

Definition at line 197 of file Machine.cpp.

CBinaryLabels * apply_locked_binary ( SGVector< index_t indices) [virtual, inherited]

Applies a locked machine on a set of indices. Error if machine is not locked. Binary case

Parameters:
indicesindex vector (of locked features) that is predicted
Returns:
resulting labels

Reimplemented from CMachine.

Definition at line 515 of file KernelMachine.cpp.

SGVector< float64_t > apply_locked_get_output ( SGVector< index_t indices) [virtual, inherited]

Applies a locked machine on a set of indices. Error if machine is not locked

Parameters:
indicesindex vector (of locked features) that is predicted
Returns:
raw output of machine

Definition at line 528 of file KernelMachine.cpp.

CLatentLabels * apply_locked_latent ( SGVector< index_t indices) [virtual, inherited]

applies a locked machine on a set of indices for latent problems

Definition at line 276 of file Machine.cpp.

CMulticlassLabels * apply_locked_multiclass ( SGVector< index_t indices) [virtual, inherited]

applies a locked machine on a set of indices for multiclass problems

Definition at line 262 of file Machine.cpp.

CRegressionLabels * apply_locked_regression ( SGVector< index_t indices) [virtual, inherited]

Applies a locked machine on a set of indices. Error if machine is not locked. Binary case

Parameters:
indicesindex vector (of locked features) that is predicted
Returns:
resulting labels

Reimplemented from CMachine.

Definition at line 521 of file KernelMachine.cpp.

CStructuredLabels * apply_locked_structured ( SGVector< index_t indices) [virtual, inherited]

applies a locked machine on a set of indices for structured problems

Definition at line 269 of file Machine.cpp.

CMulticlassLabels * apply_multiclass ( CFeatures data = NULL) [virtual, inherited]

apply machine to data in means of multiclass classification problem

Reimplemented in CMulticlassMachine, CKNN, CDistanceMachine, CVwConditionalProbabilityTree, CGaussianNaiveBayes, CConditionalProbabilityTree, CMCLDA, CQDA, CRelaxedTree, and CBaggingMachine.

Definition at line 230 of file Machine.cpp.

float64_t apply_one ( int32_t  num) [virtual, inherited]

apply kernel machine to one example

Parameters:
numwhich example to apply to
Returns:
classified value

Reimplemented from CMachine.

Definition at line 402 of file KernelMachine.cpp.

CRegressionLabels * apply_regression ( CFeatures data = NULL) [virtual, inherited]

apply kernel machine to data for regression task

Parameters:
data(test)data to be classified
Returns:
classified labels

Reimplemented from CMachine.

Definition at line 239 of file KernelMachine.cpp.

CStructuredLabels * apply_structured ( CFeatures data = NULL) [virtual, inherited]

apply machine to data in means of SO classification problem

Reimplemented in CLinearStructuredOutputMachine.

Definition at line 236 of file Machine.cpp.

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 1156 of file SGObject.cpp.

bool check_glp_status ( glp_prob *  lp) [protected]

check glpk error status

Returns:
if in good status

Definition at line 178 of file MKL.cpp.

bool cleanup_cplex ( ) [protected]

cleanup cplex

Returns:
if cleanup was successful

Definition at line 118 of file MKL.cpp.

bool cleanup_glpk ( ) [protected]

cleanup glpk

Returns:
if cleanup was successful

Definition at line 168 of file MKL.cpp.

CSGObject * clone ( ) [virtual, inherited]

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 1273 of file SGObject.cpp.

compute ElasticnetMKL dual objective

Returns:
computed dual objective

Definition at line 590 of file MKL.cpp.

compute mkl dual objective

Returns:
computed dual objective

Reimplemented in CMKLRegression.

Definition at line 1524 of file MKL.cpp.

compute mkl primal objective

Returns:
computed mkl primal objective

Definition at line 187 of file MKL.h.

float64_t compute_optimal_betas_block_norm ( float64_t beta,
const float64_t old_beta,
const int32_t  num_kernels,
const float64_t sumw,
const float64_t  suma,
const float64_t  mkl_objective 
) [protected]

given the alphas, compute the corresponding optimal betas

Parameters:
betanew betas (kernel weights)
old_betaold betas (previous kernel weights)
num_kernelsnumber of kernels
sumw1/2*alpha'*K_j*alpha for each kernel j
suma(sum over alphas)
mkl_objectivethe current mkl objective
Returns:
new objective value

Definition at line 665 of file MKL.cpp.

float64_t compute_optimal_betas_directly ( float64_t beta,
const float64_t old_beta,
const int32_t  num_kernels,
const float64_t sumw,
const float64_t  suma,
const float64_t  mkl_objective 
) [protected]

given the alphas, compute the corresponding optimal betas

Parameters:
betanew betas (kernel weights)
old_betaold betas (previous kernel weights)
num_kernelsnumber of kernels
sumw1/2*alpha'*K_j*alpha for each kernel j
suma(sum over alphas)
mkl_objectivethe current mkl objective
Returns:
new objective value

Definition at line 701 of file MKL.cpp.

float64_t compute_optimal_betas_elasticnet ( float64_t beta,
const float64_t old_beta,
const int32_t  num_kernels,
const float64_t sumw,
const float64_t  suma,
const float64_t  mkl_objective 
) [protected]

given the alphas, compute the corresponding optimal betas

Parameters:
betanew betas (kernel weights)
old_betaold betas (previous kernel weights)
num_kernelsnumber of kernels
sumw1/2*alpha'*K_j*alpha for each kernel j
suma(sum over alphas)
mkl_objectivethe current mkl objective
Returns:
new objective value

Definition at line 471 of file MKL.cpp.

float64_t compute_optimal_betas_newton ( float64_t beta,
const float64_t old_beta,
int32_t  num_kernels,
const float64_t sumw,
float64_t  suma,
float64_t  mkl_objective 
) [protected]

given the alphas, compute the corresponding optimal betas

Parameters:
betanew betas (kernel weights)
old_betaold betas (previous kernel weights)
num_kernelsnumber of kernels
sumw1/2*alpha'*K_j*alpha for each kernel j
suma(sum over alphas)
mkl_objectivethe current mkl objective
Returns:
new objective value

Definition at line 790 of file MKL.cpp.

float64_t compute_optimal_betas_via_cplex ( float64_t beta,
const float64_t old_beta,
int32_t  num_kernels,
const float64_t sumw,
float64_t  suma,
int32_t &  inner_iters 
) [protected]

given the alphas, compute the corresponding optimal betas using a lp for 1-norm mkl, a qcqp for 2-norm mkl and an iterated qcqp for general q-norm mkl.

Parameters:
betanew betas (kernel weights)
old_betaold betas (previous kernel weights)
num_kernelsnumber of kernels
sumw1/2*alpha'*K_j*alpha for each kernel j
suma(sum over alphas)
inner_itersnumber of internal iterations (for statistics)
Returns:
new objective value

Definition at line 982 of file MKL.cpp.

float64_t compute_optimal_betas_via_glpk ( float64_t beta,
const float64_t old_beta,
int  num_kernels,
const float64_t sumw,
float64_t  suma,
int32_t &  inner_iters 
) [protected]

given the alphas, compute the corresponding optimal betas using a lp for 1-norm mkl

Parameters:
betanew betas (kernel weights)
old_betaold betas (previous kernel weights)
num_kernelsnumber of kernels
sumw1/2*alpha'*K_j*alpha for each kernel j
suma(sum over alphas)
inner_itersnumber of internal iterations (for statistics)
Returns:
new objective value

Definition at line 1325 of file MKL.cpp.

virtual float64_t compute_sum_alpha ( ) [pure virtual]

compute beta independent term from objective, e.g., in 2-class MKL sum_i alpha_i etc

Implemented in CMKLRegression, CMKLClassification, and CMKLOneClass.

void compute_sum_beta ( float64_t sumw) [virtual]

compute 1/2*alpha'*K_j*alpha for each kernel j (beta dependent term from objective)

Parameters:
sumwvector of size num_kernels to hold the result

Definition at line 1479 of file MKL.cpp.

compute svm dual objective

Returns:
computed dual objective

Definition at line 242 of file SVM.cpp.

compute svm primal objective

Returns:
computed svm primal objective

Definition at line 267 of file SVM.cpp.

virtual bool converged ( ) [protected, virtual]

check if mkl converged, i.e. 'gap' is below epsilon

Returns:
whether mkl converged

Definition at line 404 of file MKL.h.

bool create_new_model ( int32_t  num) [inherited]

create new model

Parameters:
numnumber of alphas and support vectors in new model

Definition at line 191 of file KernelMachine.cpp.

void data_lock ( CLabels labs,
CFeatures features = NULL 
) [virtual, inherited]

Locks the machine on given labels and data. After this call, only train_locked and apply_locked may be called.

Computes kernel matrix to speed up train/apply calls

Parameters:
labslabels used for locking
featuresfeatures used for locking

Reimplemented from CMachine.

Definition at line 620 of file KernelMachine.cpp.

void data_unlock ( ) [virtual, inherited]

Unlocks a locked machine and restores previous state

Reimplemented from CMachine.

Definition at line 649 of file KernelMachine.cpp.

virtual CSGObject* deep_copy ( ) const [virtual, inherited]

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

Definition at line 126 of file SGObject.h.

void elasticnet_dual ( float64_t ff,
float64_t gg,
float64_t hh,
const float64_t del,
const float64_t nm,
int32_t  len,
const float64_t lambda 
) [protected]

helper function to compute the elastic-net objective

Definition at line 563 of file MKL.cpp.

void elasticnet_transform ( float64_t beta,
float64_t  lmd,
int32_t  len 
) [protected]

helper function to compute the elastic-net sub-kernel weights

Definition at line 345 of file MKL.h.

bool equals ( CSGObject other,
float64_t  accuracy = 0.0 
) [virtual, inherited]

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)
Returns:
true if all parameters were equal, false if not

Definition at line 1177 of file SGObject.cpp.

float64_t get_alpha ( int32_t  idx) [inherited]

get alpha at given index

Parameters:
idxindex of alpha
Returns:
alpha

Definition at line 137 of file KernelMachine.cpp.

SGVector< float64_t > get_alphas ( ) [inherited]
Returns:
vector of alphas

Definition at line 186 of file KernelMachine.cpp.

bool get_batch_computation_enabled ( ) [inherited]

check if batch computation is enabled

Returns:
if batch computation is enabled

Definition at line 96 of file KernelMachine.cpp.

float64_t get_bias ( ) [inherited]

get bias

Returns:
bias

Definition at line 121 of file KernelMachine.cpp.

bool get_bias_enabled ( ) [inherited]

get state of bias

Returns:
state of bias

Definition at line 116 of file KernelMachine.cpp.

float64_t get_C1 ( ) [inherited]

get C1

Returns:
C1

Definition at line 159 of file SVM.h.

float64_t get_C2 ( ) [inherited]

get C2

Returns:
C2

Definition at line 165 of file SVM.h.

EMachineType get_classifier_type ( ) [virtual, inherited]
float64_t get_epsilon ( ) [inherited]

get epsilon

Returns:
epsilon

Definition at line 147 of file SVM.h.

SGIO * get_global_io ( ) [inherited]

get the io object

Returns:
io object

Definition at line 174 of file SGObject.cpp.

Parallel * get_global_parallel ( ) [inherited]

get the parallel object

Returns:
parallel object

Definition at line 209 of file SGObject.cpp.

Version * get_global_version ( ) [inherited]

get the version object

Returns:
version object

Definition at line 222 of file SGObject.cpp.

get state of optimization (interleaved or wrapper)

Returns:
true if interleaved optimization is used; wrapper otherwise

Definition at line 178 of file MKL.h.

CKernel * get_kernel ( ) [inherited]

get kernel

Returns:
kernel

Definition at line 85 of file KernelMachine.cpp.

CLabels * get_labels ( ) [virtual, inherited]

get labels

Returns:
labels

Definition at line 86 of file Machine.cpp.

bool get_linadd_enabled ( ) [inherited]

check if linadd is enabled

Returns:
if linadd is enabled

Definition at line 106 of file KernelMachine.cpp.

SGVector< float64_t > get_linear_term ( ) [virtual, inherited]

get linear term

Returns:
the linear term

Definition at line 332 of file SVM.cpp.

float64_t * get_linear_term_array ( ) [protected, virtual, inherited]

get linear term copy as dynamic array

Returns:
linear term copied to a dynamic array

Definition at line 302 of file SVM.cpp.

virtual EProblemType get_machine_problem_type ( ) const [virtual, inherited]

returns type of problem machine solves

Reimplemented in CBaseMulticlassMachine.

Definition at line 292 of file Machine.h.

float64_t get_max_train_time ( ) [inherited]

get maximum training time

Returns:
maximum training time

Definition at line 97 of file Machine.cpp.

get mkl epsilon for weights (optimization accuracy for kernel weights)

Returns:
epsilon for weights

Definition at line 215 of file MKL.h.

int32_t get_mkl_iterations ( )

get number of MKL iterations

Returns:
mkl_iterations

Definition at line 221 of file MKL.h.

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

Definition at line 1060 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 1084 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 1097 of file SGObject.cpp.

virtual const char* get_name ( ) const [virtual]
Returns:
object name

Reimplemented from CSVM.

Reimplemented in CMKLRegression, CMKLClassification, and CMKLOneClass.

Definition at line 260 of file MKL.h.

float64_t get_nu ( ) [inherited]

get nu

Returns:
nu

Definition at line 153 of file SVM.h.

int32_t get_num_support_vectors ( ) [inherited]

get number of support vectors

Returns:
number of support vectors

Definition at line 166 of file KernelMachine.cpp.

float64_t get_objective ( ) [inherited]

get objective

Returns:
objective

Definition at line 216 of file SVM.h.

int32_t get_qpsize ( ) [inherited]

get qpsize

Returns:
qpsize

Definition at line 171 of file SVM.h.

bool get_shrinking_enabled ( ) [inherited]

get state of shrinking

Returns:
if shrinking is enabled

Definition at line 186 of file SVM.h.

ESolverType get_solver_type ( ) [inherited]

get solver type

Returns:
solver

Definition at line 112 of file Machine.cpp.

int32_t get_support_vector ( int32_t  idx) [inherited]

get support vector at given index

Parameters:
idxindex of support vector
Returns:
support vector

Definition at line 131 of file KernelMachine.cpp.

SGVector< int32_t > get_support_vectors ( ) [inherited]
Returns:
all support vectors

Definition at line 181 of file KernelMachine.cpp.

CSVM* get_svm ( )

get SVM that is used as constraint generator in MKL SIP

Returns:
svm

Definition at line 132 of file MKL.h.

float64_t get_tube_epsilon ( ) [inherited]

get tube epsilon

Returns:
tube epsilon

Definition at line 135 of file SVM.h.

bool init_cplex ( ) [protected]

init cplex

Returns:
if init was successful

Definition at line 69 of file MKL.cpp.

bool init_glpk ( ) [protected]

init glpk

Returns:
if init was successful

Definition at line 154 of file MKL.cpp.

bool init_kernel_optimization ( ) [inherited]

initialise kernel optimisation

Returns:
if operation was successful

Definition at line 208 of file KernelMachine.cpp.

void init_solver ( ) [protected]

initialize solver such as glpk or cplex

Definition at line 51 of file MKL.cpp.

virtual void init_training ( ) [protected, pure virtual]

check run before starting training (to e.g. check if labeling is two-class labeling in classification case

Implemented in CMKLRegression, CMKLClassification, and CMKLOneClass.

bool is_data_locked ( ) const [inherited]
Returns:
whether this machine is locked

Definition at line 289 of file Machine.h.

bool is_generic ( EPrimitiveType *  generic) const [virtual, inherited]

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 228 of file SGObject.cpp.

virtual bool is_label_valid ( CLabels lab) const [protected, virtual, inherited]

check whether the labels is valid.

Subclasses can override this to implement their check of label types.

Parameters:
labthe labels being checked, guaranteed to be non-NULL

Reimplemented in CGaussianProcessBinaryClassification, CGaussianProcessRegression, and CBaseMulticlassMachine.

Definition at line 341 of file Machine.h.

bool load ( FILE *  svm_file) [inherited]

load a SVM from file

Parameters:
svm_filethe file handle

Definition at line 90 of file SVM.cpp.

DynArray< TParameter * > * load_all_file_parameters ( int32_t  file_version,
int32_t  current_version,
CSerializableFile file,
const char *  prefix = "" 
) [inherited]

maps all parameters of this instance to the provided file version and loads all parameter data from the file into an array, which is sorted (basically calls load_file_parameter(...) for all parameters and puts all results into a sorted array)

Parameters:
file_versionparameter version of the file
current_versionversion from which mapping begins (you want to use Version::get_version_parameter() for this in most cases)
filefile to load from
prefixprefix for members
Returns:
(sorted) array of created TParameter instances with file data

Definition at line 633 of file SGObject.cpp.

DynArray< TParameter * > * load_file_parameters ( const SGParamInfo param_info,
int32_t  file_version,
CSerializableFile file,
const char *  prefix = "" 
) [inherited]

loads some specified parameters from a file with a specified version The provided parameter info has a version which is recursively mapped until the file parameter version is reached. Note that there may be possibly multiple parameters in the mapping, therefore, a set of TParameter instances is returned

Parameters:
param_infoinformation of parameter
file_versionparameter version of the file, must be <= provided parameter version
filefile to load from
prefixprefix for members
Returns:
new array with TParameter instances with the attached data

Definition at line 474 of file SGObject.cpp.

bool load_serializable ( CSerializableFile file,
const char *  prefix = "",
int32_t  param_version = Version::get_version_parameter() 
) [virtual, inherited]

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
param_version(optional) a parameter version different to (this is mainly for testing, better do not use)
Returns:
TRUE if done, otherwise FALSE

Definition at line 305 of file SGObject.cpp.

void load_serializable_post ( ) throw (ShogunException) [protected, virtual, inherited]

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 occurres.

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

Definition at line 989 of file SGObject.cpp.

void load_serializable_pre ( ) throw (ShogunException) [protected, virtual, inherited]

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 occurres.

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

Definition at line 984 of file SGObject.cpp.

MACHINE_PROBLEM_TYPE ( PT_BINARY  ) [inherited]

problem type

void map_parameters ( DynArray< TParameter * > *  param_base,
int32_t &  base_version,
DynArray< const SGParamInfo * > *  target_param_infos 
) [inherited]

Takes a set of TParameter instances (base) with a certain version and a set of target parameter infos and recursively maps the base level wise to the current version using CSGObject::migrate(...). The base is replaced. After this call, the base version containing parameters should be of same version/type as the initial target parameter infos. Note for this to work, the migrate methods and all the internal parameter mappings have to match

Parameters:
param_baseset of TParameter instances that are mapped to the provided target parameter infos
base_versionversion of the parameter base
target_param_infosset of SGParamInfo instances that specify the target parameter base

Definition at line 671 of file SGObject.cpp.

TParameter * migrate ( DynArray< TParameter * > *  param_base,
const SGParamInfo target 
) [protected, virtual, inherited]

creates a new TParameter instance, which contains migrated data from the version that is provided. The provided parameter data base is used for migration, this base is a collection of all parameter data of the previous version. Migration is done FROM the data in param_base TO the provided param info Migration is always one version step. Method has to be implemented in subclasses, if no match is found, base method has to be called.

If there is an element in the param_base which equals the target, a copy of the element is returned. This represents the case when nothing has changed and therefore, the migrate method is not overloaded in a subclass

Parameters:
param_baseset of TParameter instances to use for migration
targetparameter info for the resulting TParameter
Returns:
a new TParameter instance with migrated data from the base of the type which is specified by the target parameter

Definition at line 878 of file SGObject.cpp.

void one_to_one_migration_prepare ( DynArray< TParameter * > *  param_base,
const SGParamInfo target,
TParameter *&  replacement,
TParameter *&  to_migrate,
char *  old_name = NULL 
) [protected, virtual, inherited]

This method prepares everything for a one-to-one parameter migration. One to one here means that only ONE element of the parameter base is needed for the migration (the one with the same name as the target). Data is allocated for the target (in the type as provided in the target SGParamInfo), and a corresponding new TParameter instance is written to replacement. The to_migrate pointer points to the single needed TParameter instance needed for migration. If a name change happened, the old name may be specified by old_name. In addition, the m_delete_data flag of to_migrate is set to true. So if you want to migrate data, the only thing to do after this call is converting the data in the m_parameter fields. If unsure how to use - have a look into an example for this. (base_migration_type_conversion.cpp for example)

Parameters:
param_baseset of TParameter instances to use for migration
targetparameter info for the resulting TParameter
replacement(used as output) here the TParameter instance which is returned by migration is created into
to_migratethe only source that is used for migration
old_namewith this parameter, a name change may be specified

Definition at line 818 of file SGObject.cpp.

bool perform_mkl_step ( const float64_t sumw,
float64_t  suma 
) [virtual]

perform single mkl iteration

given sum of alphas, objectives for current alphas for each kernel and current kernel weighting compute the corresponding optimal kernel weighting (all via get/set_subkernel_weights in CCombinedKernel)

Parameters:
sumwvector of 1/2*alpha'*K_j*alpha for each kernel j
sumascalar sum_i alpha_i etc.

Definition at line 402 of file MKL.cpp.

void perform_mkl_step ( float64_t beta,
float64_t old_beta,
int  num_kernels,
int32_t *  label,
int32_t *  active2dnum,
float64_t a,
float64_t lin,
float64_t sumw,
int32_t &  inner_iters 
) [protected]

perform single mkl iteration

given the alphas, compute the corresponding optimal betas

Parameters:
betanew betas (kernel weights)
old_betaold betas (previous kernel weights)
num_kernelsnumber of kernels
label(from svmlight label)
active2dnum(from svmlight active2dnum)
a(from svmlight alphas)
lin(from svmlight linear components)
sumw1/2*alpha'*K_j*alpha for each kernel j
inner_itersnumber of required internal iterations
static bool perform_mkl_step_helper ( CMKL mkl,
const float64_t sumw,
const float64_t  suma 
) [static]

callback helper function calling perform_mkl_step

Parameters:
mklMKL object
sumwvector of 1/2*alpha'*K_j*alpha for each kernel j
sumascalar sum_i alpha_i etc.

Definition at line 241 of file MKL.h.

virtual void post_lock ( CLabels labs,
CFeatures features 
) [virtual, inherited]

post lock

Reimplemented in CMultitaskLinearMachine.

Definition at line 280 of file Machine.h.

void print_modsel_params ( ) [inherited]

prints all parameter registered for model selection and their type

Definition at line 1036 of file SGObject.cpp.

void print_serializable ( const char *  prefix = "") [virtual, inherited]

prints registered parameters out

Parameters:
prefixprefix for members

Definition at line 240 of file SGObject.cpp.

bool save ( FILE *  svm_file) [inherited]

write a SVM to a file

Parameters:
svm_filethe file handle

Definition at line 206 of file SVM.cpp.

bool save_serializable ( CSerializableFile file,
const char *  prefix = "",
int32_t  param_version = Version::get_version_parameter() 
) [virtual, inherited]

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
param_version(optional) a parameter version different to (this is mainly for testing, better do not use)
Returns:
TRUE if done, otherwise FALSE

Definition at line 246 of file SGObject.cpp.

void save_serializable_post ( ) throw (ShogunException) [protected, virtual, inherited]

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 occurres.

Reimplemented in CKernel.

Definition at line 999 of file SGObject.cpp.

void save_serializable_pre ( ) throw (ShogunException) [protected, virtual, inherited]

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 occurres.

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

Definition at line 994 of file SGObject.cpp.

bool set_alpha ( int32_t  idx,
float64_t  val 
) [inherited]

set alpha at given index to given value

Parameters:
idxindex of alpha vector
valnew value of alpha vector
Returns:
if operation was successful

Definition at line 156 of file KernelMachine.cpp.

void set_alphas ( SGVector< float64_t alphas) [inherited]

set alphas to given values

Parameters:
alphasfloat vector with all alphas to set

Definition at line 171 of file KernelMachine.cpp.

void set_batch_computation_enabled ( bool  enable) [inherited]

set batch computation enabled

Parameters:
enableif batch computation shall be enabled

Definition at line 91 of file KernelMachine.cpp.

void set_bias ( float64_t  bias) [inherited]

set bias to given value

Parameters:
biasnew bias

Definition at line 126 of file KernelMachine.cpp.

void set_bias_enabled ( bool  enable_bias) [inherited]

set state of bias

Parameters:
enable_biasif bias shall be enabled

Definition at line 111 of file KernelMachine.cpp.

void set_C ( float64_t  c_neg,
float64_t  c_pos 
) [inherited]

set C

Parameters:
c_negnew C constant for negatively labeled examples
c_posnew C constant for positively labeled examples

Note that not all SVMs support this (however at least CLibSVM and CSVMLight do)

Definition at line 116 of file SVM.h.

void set_C_mkl ( float64_t  C)

set C mkl

Parameters:
Cnew C_mkl

Definition at line 142 of file MKL.h.

void set_callback_function ( CMKL m,
bool(*)(CMKL *mkl, const float64_t *sumw, const float64_t suma)  cb 
) [inherited]

set callback function svm optimizers may call when they have a new (small) set of alphas

Parameters:
mpointer to mkl object
cbcallback function

Definition at line 232 of file SVM.cpp.

SVM to use as constraint generator in MKL SIP

Parameters:
ssvm

Definition at line 112 of file MKL.h.

void set_defaults ( int32_t  num_sv = 0) [inherited]

set default values for members a SVM object

Definition at line 48 of file SVM.cpp.

void set_elasticnet_lambda ( float64_t  elasticnet_lambda)

set elasticnet lambda

Parameters:
elasticnet_lambdanew elastic net lambda (must be 0<=lambda<=1) lambda=0: L1-MKL lambda=1: Linfinity-MKL

Definition at line 381 of file MKL.cpp.

void set_epsilon ( float64_t  eps) [inherited]

set epsilon

Parameters:
epsnew epsilon

Definition at line 123 of file SVM.h.

void set_generic< complex128_t > ( ) [inherited]

set generic type to T

Definition at line 41 of file SGObject.cpp.

void set_global_io ( SGIO io) [inherited]

set the io object

Parameters:
ioio object to use

Definition at line 167 of file SGObject.cpp.

void set_global_parallel ( Parallel parallel) [inherited]

set the parallel object

Parameters:
parallelparallel object to use

Definition at line 180 of file SGObject.cpp.

void set_global_version ( Version version) [inherited]

set the version object

Parameters:
versionversion object to use

Definition at line 215 of file SGObject.cpp.

void set_interleaved_optimization_enabled ( bool  enable)

set state of optimization (interleaved or wrapper)

Parameters:
enableif true interleaved optimization is used; wrapper otherwise

Definition at line 169 of file MKL.h.

void set_kernel ( CKernel k) [inherited]

set kernel

Parameters:
kkernel

Definition at line 78 of file KernelMachine.cpp.

void set_labels ( CLabels lab) [virtual, inherited]

set labels

Parameters:
lablabels

Reimplemented in CGaussianProcessMachine, CStructuredOutputMachine, CRelaxedTree, and CMulticlassMachine.

Definition at line 75 of file Machine.cpp.

void set_linadd_enabled ( bool  enable) [inherited]

set linadd enabled

Parameters:
enableif linadd shall be enabled

Definition at line 101 of file KernelMachine.cpp.

void set_linear_term ( const SGVector< float64_t linear_term) [virtual, inherited]

set linear term of the QP

Parameters:
linear_termthe linear term

Definition at line 314 of file SVM.cpp.

void set_max_train_time ( float64_t  t) [inherited]

set maximum training time

Parameters:
tmaximimum training time

Definition at line 92 of file Machine.cpp.

set block norm q (used in block norm mkl)

Parameters:
qmixed norm (1<=q<=inf)

Definition at line 394 of file MKL.cpp.

void set_mkl_epsilon ( float64_t  eps)

set mkl epsilon (optimization accuracy for kernel weights)

Parameters:
epsnew weight_epsilon

Definition at line 209 of file MKL.h.

void set_mkl_norm ( float64_t  norm)

set mkl norm

Parameters:
normnew mkl norm (must be greater equal 1)

Definition at line 372 of file MKL.cpp.

void set_nu ( float64_t  nue) [inherited]

set nu

Parameters:
nuenew nu

Definition at line 105 of file SVM.h.

void set_objective ( float64_t  v) [inherited]

set objective

Parameters:
vobjective

Definition at line 207 of file SVM.h.

void set_qnorm_constraints ( float64_t beta,
int32_t  num_kernels 
) [protected]

set qnorm mkl constraints

Definition at line 1574 of file MKL.cpp.

void set_qpsize ( int32_t  qps) [inherited]

set qpsize

Parameters:
qpsnew qpsize

Definition at line 141 of file SVM.h.

void set_shrinking_enabled ( bool  enable) [inherited]

set state of shrinking

Parameters:
enableif shrinking will be enabled

Definition at line 177 of file SVM.h.

void set_solver_type ( ESolverType  st) [inherited]

set solver type

Parameters:
stsolver type

Definition at line 107 of file Machine.cpp.

void set_store_model_features ( bool  store_model) [virtual, inherited]

Setter for store-model-features-after-training flag

Parameters:
store_modelwhether model should be stored after training

Definition at line 117 of file Machine.cpp.

bool set_support_vector ( int32_t  idx,
int32_t  val 
) [inherited]

set support vector at given index to given value

Parameters:
idxindex of support vector
valnew value of support vector
Returns:
if operation was successful

Definition at line 146 of file KernelMachine.cpp.

void set_support_vectors ( SGVector< int32_t >  svs) [inherited]

set support vectors to given values

Parameters:
svsinteger vector with all support vectors indexes to set

Definition at line 176 of file KernelMachine.cpp.

void set_svm ( CSVM s)

SVM to use as constraint generator in MKL SIP

Parameters:
ssvm

Definition at line 121 of file MKL.h.

void set_tube_epsilon ( float64_t  eps) [inherited]

set tube epsilon

Parameters:
epsnew tube epsilon

Definition at line 129 of file SVM.h.

virtual CSGObject* shallow_copy ( ) const [virtual, inherited]

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

Reimplemented in CGaussianKernel.

Definition at line 117 of file SGObject.h.

void store_model_features ( ) [protected, virtual, inherited]

Stores feature data of the SV indices and sets it to the lhs of the underlying kernel. Then, all SV indices are set to identity.

May be overwritten by subclasses in case the model should be stored differently.

Reimplemented from CMachine.

Definition at line 450 of file KernelMachine.cpp.

bool supports_locking ( ) const [virtual, inherited]
Returns:
whether machine supports locking

Reimplemented from CMachine.

Definition at line 707 of file KernelMachine.cpp.

bool train ( CFeatures data = NULL) [virtual, inherited]

train machine

Parameters:
datatraining data (parameter can be avoided if distance or kernel-based classifiers are used and distance/kernels are initialized with train data). If flag is set, model features will be stored after training.
Returns:
whether training was successful

Reimplemented in CRelaxedTree, CSGDQN, and COnlineSVMSGD.

Definition at line 49 of file Machine.cpp.

bool train_locked ( SGVector< index_t indices) [virtual, inherited]

Trains a locked machine on a set of indices. Error if machine is not locked

Parameters:
indicesindex vector (of locked features) that is used for training
Returns:
whether training was successful

Reimplemented from CMachine.

Definition at line 479 of file KernelMachine.cpp.

bool train_machine ( CFeatures data = NULL) [protected, virtual]

train MKL classifier

Parameters:
datatraining data (parameter can be avoided if distance or kernel-based classifiers are used and distance/kernels are initialized with train data)
Returns:
whether training was successful

Reimplemented from CMachine.

Definition at line 196 of file MKL.cpp.

virtual bool train_require_labels ( ) const [protected, virtual, inherited]

returns whether machine require labels for training

Reimplemented in COnlineLinearMachine, CHierarchical, CLinearLatentMachine, CVwConditionalProbabilityTree, CConditionalProbabilityTree, and CLibSVMOneClass.

Definition at line 347 of file Machine.h.

void unset_generic ( ) [inherited]

unset generic type

this has to be called in classes specializing a template class

Definition at line 235 of file SGObject.cpp.

bool update_parameter_hash ( ) [virtual, inherited]

Updates the hash of current parameter combination.

Returns:
bool if parameter combination has changed since last update.

Definition at line 187 of file SGObject.cpp.


Friends And Related Function Documentation

friend class CMulticlassSVM [friend, inherited]

Definition at line 272 of file SVM.h.


Member Data Documentation

float64_t* beta_local [protected]

sub-kernel weights on the L1-term of ElasticnetMKL

Definition at line 468 of file MKL.h.

float64_t C1 [protected, inherited]

C1 regularization const

Definition at line 255 of file SVM.h.

float64_t C2 [protected, inherited]

C2

Definition at line 257 of file SVM.h.

float64_t C_mkl [protected]

C_mkl

Definition at line 453 of file MKL.h.

bool(* callback)(CMKL *mkl, const float64_t *sumw, const float64_t suma) [protected, inherited]

callback function svm optimizers may call when they have a new (small) set of alphas

Definition at line 267 of file SVM.h.

float64_t ent_lambda [protected]

Sparsity trade-off parameter used in ElasticnetMKL must be 0<=lambda<=1 lambda=0: L1-MKL lambda=1: Linfinity-MKL

Definition at line 461 of file MKL.h.

CPXENVptr env [protected]

env

Definition at line 489 of file MKL.h.

float64_t epsilon [protected, inherited]

epsilon

Definition at line 249 of file SVM.h.

bool interleaved_optimization [protected]

whether to use mkl wrapper or interleaved opt.

Definition at line 474 of file MKL.h.

SGIO* io [inherited]

io

Definition at line 473 of file SGObject.h.

CKernel* kernel [protected, inherited]

kernel

Definition at line 310 of file KernelMachine.h.

CPXLPptr lp_cplex [protected]

lp

Definition at line 491 of file MKL.h.

glp_prob* lp_glpk [protected]

lp

Definition at line 496 of file MKL.h.

glp_smcp* lp_glpk_parm [protected]

lp parameters

Definition at line 499 of file MKL.h.

bool lp_initialized [protected]

if lp is initialized

Definition at line 502 of file MKL.h.

SGVector<float64_t> m_alpha [protected, inherited]

coefficients alpha

Definition at line 331 of file KernelMachine.h.

float64_t m_bias [protected, inherited]

bias term b

Definition at line 328 of file KernelMachine.h.

CCustomKernel* m_custom_kernel [protected, inherited]

is filled with pre-computed custom kernel on data lock

Definition at line 313 of file KernelMachine.h.

bool m_data_locked [protected, inherited]

whether data is locked

Definition at line 363 of file Machine.h.

parameters wrt which we can compute gradients

Definition at line 488 of file SGObject.h.

uint32_t m_hash [inherited]

Hash of parameter values

Definition at line 494 of file SGObject.h.

CKernel* m_kernel_backup [protected, inherited]

old kernel is stored here on data lock

Definition at line 316 of file KernelMachine.h.

CLabels* m_labels [protected, inherited]

labels

Definition at line 354 of file Machine.h.

SGVector<float64_t> m_linear_term [protected, inherited]

linear term in qp

Definition at line 244 of file SVM.h.

float64_t m_max_train_time [protected, inherited]

maximum training time

Definition at line 351 of file Machine.h.

model selection parameters

Definition at line 485 of file SGObject.h.

map for different parameter versions

Definition at line 491 of file SGObject.h.

Parameter* m_parameters [inherited]

parameters

Definition at line 482 of file SGObject.h.

ESolverType m_solver_type [protected, inherited]

solver type

Definition at line 357 of file Machine.h.

bool m_store_model_features [protected, inherited]

whether model features should be stored after training

Definition at line 360 of file Machine.h.

SGVector<int32_t> m_svs [protected, inherited]

array of ``support vectors'' (indices of feature objects)

Definition at line 334 of file KernelMachine.h.

CMKL* mkl [protected, inherited]

mkl object that svm optimizers need to pass when calling the callback function

Definition at line 270 of file SVM.h.

Sparsity trade-off parameter used in block norm MKL should be 1 <= mkl_block_norm <= inf

Definition at line 465 of file MKL.h.

float64_t mkl_epsilon [protected]

mkl_epsilon for multiple kernel learning

Definition at line 472 of file MKL.h.

int32_t mkl_iterations [protected]

number of mkl steps

Definition at line 470 of file MKL.h.

float64_t mkl_norm [protected]

norm used in mkl must be > 0

Definition at line 455 of file MKL.h.

float64_t nu [protected, inherited]

nu

Definition at line 253 of file SVM.h.

float64_t objective [protected, inherited]

objective

Definition at line 259 of file SVM.h.

Parallel* parallel [inherited]

parallel

Definition at line 476 of file SGObject.h.

int32_t qpsize [protected, inherited]

qpsize

Definition at line 261 of file SVM.h.

float64_t rho [protected]

objective after mkl iterations

Definition at line 482 of file MKL.h.

CSVM* svm [protected]

wrapper SVM

Definition at line 451 of file MKL.h.

bool svm_loaded [protected, inherited]

if SVM is loaded

Definition at line 247 of file SVM.h.

measures training time for use with get_max_train_time()

Definition at line 485 of file MKL.h.

float64_t tube_epsilon [protected, inherited]

tube epsilon for support vector regression

Definition at line 251 of file SVM.h.

bool use_batch_computation [protected, inherited]

if batch computation is enabled

Definition at line 319 of file KernelMachine.h.

bool use_bias [protected, inherited]

if bias shall be used

Definition at line 325 of file KernelMachine.h.

bool use_linadd [protected, inherited]

if linadd is enabled

Definition at line 322 of file KernelMachine.h.

bool use_shrinking [protected, inherited]

if shrinking shall be used

Definition at line 263 of file SVM.h.

Version* version [inherited]

version

Definition at line 479 of file SGObject.h.

float64_t* W [protected]

partial objectives (one per kernel)

Definition at line 477 of file MKL.h.

float64_t w_gap [protected]

gap between iterations

Definition at line 480 of file MKL.h.


The documentation for this class was generated from the following files:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines

SHOGUN Machine Learning Toolbox - Documentation