SHOGUN
4.0.0
|
Class CStochasticSOSVM solves SOSVM using stochastic subgradient descent on the SVM primal problem [1], which is equivalent to SGD or Pegasos [2]. This class is inspired by the matlab SGD implementation in [3].
[1] N. Ratliff, J. A. Bagnell, and M. Zinkevich. (online) subgradient methods for structured prediction. AISTATS, 2007. [2] S. Shalev-Shwartz, Y. Singer, N. Srebro. Pegasos: Primal Estimated sub-GrAdient SOlver for SVM. ICML 2007. [3] S. Lacoste-Julien, M. Jaggi, M. Schmidt and P. Pletscher. Block-Coordinate Frank-Wolfe Optimization for Structural SVMs. ICML 2013.
Definition at line 32 of file StochasticSOSVM.h.
Public Attributes | |
SGIO * | io |
Parallel * | parallel |
Version * | version |
Parameter * | m_parameters |
Parameter * | m_model_selection_parameters |
Parameter * | m_gradient_parameters |
ParameterMap * | m_parameter_map |
uint32_t | m_hash |
Protected Attributes | |
SGVector< float64_t > | m_w |
CStructuredModel * | m_model |
CLossFunction * | m_surrogate_loss |
CSOSVMHelper * | m_helper |
bool | m_verbose |
float64_t | m_max_train_time |
CLabels * | m_labels |
ESolverType | m_solver_type |
bool | m_store_model_features |
bool | m_data_locked |
CStochasticSOSVM | ( | ) |
default constructor
Definition at line 18 of file StochasticSOSVM.cpp.
CStochasticSOSVM | ( | CStructuredModel * | model, |
CStructuredLabels * | labs, | ||
bool | do_weighted_averaging = true , |
||
bool | verbose = false |
||
) |
standard constructor
model | structured model with application specific functions |
labs | structured labels |
do_weighted_averaging | whether mix w with previous average weights |
verbose | whether compute debug information, such as primal value, duality gap etc. |
Definition at line 24 of file StochasticSOSVM.cpp.
~CStochasticSOSVM | ( | ) |
destructor
Definition at line 58 of file StochasticSOSVM.cpp.
apply machine to data if data is not specified apply to the current features
data | (test)data to be classified |
Definition at line 160 of file Machine.cpp.
|
virtualinherited |
apply machine to data in means of binary classification problem
Reimplemented in CKernelMachine, COnlineLinearMachine, CWDSVMOcas, CNeuralNetwork, CLinearMachine, CGaussianProcessClassification, CDomainAdaptationSVMLinear, CPluginEstimate, and CBaggingMachine.
Definition at line 216 of file Machine.cpp.
|
virtualinherited |
apply machine to data in means of latent problem
Reimplemented in CLinearLatentMachine.
Definition at line 240 of file Machine.cpp.
Applies a locked machine on a set of indices. Error if machine is not locked
indices | index vector (of locked features) that is predicted |
Definition at line 195 of file Machine.cpp.
|
virtualinherited |
applies a locked machine on a set of indices for binary problems
Reimplemented in CKernelMachine, and CMultitaskLinearMachine.
Definition at line 246 of file Machine.cpp.
|
virtualinherited |
applies a locked machine on a set of indices for latent problems
Definition at line 274 of file Machine.cpp.
|
virtualinherited |
applies a locked machine on a set of indices for multiclass problems
Definition at line 260 of file Machine.cpp.
|
virtualinherited |
applies a locked machine on a set of indices for regression problems
Reimplemented in CKernelMachine.
Definition at line 253 of file Machine.cpp.
|
virtualinherited |
applies a locked machine on a set of indices for structured problems
Definition at line 267 of file Machine.cpp.
|
virtualinherited |
apply machine to data in means of multiclass classification problem
Reimplemented in CNeuralNetwork, CCHAIDTree, CCARTree, CGaussianProcessClassification, CMulticlassMachine, CKNN, CC45ClassifierTree, CID3ClassifierTree, CDistanceMachine, CVwConditionalProbabilityTree, CGaussianNaiveBayes, CConditionalProbabilityTree, CMCLDA, CQDA, CRelaxedTree, and CBaggingMachine.
Definition at line 228 of file Machine.cpp.
|
virtualinherited |
applies to one vector
Reimplemented in CKernelMachine, CRelaxedTree, CWDSVMOcas, COnlineLinearMachine, CLinearMachine, CMultitaskLinearMachine, CMulticlassMachine, CKNN, CDistanceMachine, CMultitaskLogisticRegression, CMultitaskLeastSquaresRegression, CScatterSVM, CGaussianNaiveBayes, CPluginEstimate, and CFeatureBlockLogisticRegression.
|
virtualinherited |
apply machine to data in means of regression problem
Reimplemented in CKernelMachine, CWDSVMOcas, COnlineLinearMachine, CNeuralNetwork, CCHAIDTree, CStochasticGBMachine, CCARTree, CLinearMachine, CGaussianProcessRegression, and CBaggingMachine.
Definition at line 222 of file Machine.cpp.
|
virtualinherited |
apply structured machine to data for Structured Output (SO) problem
data | (test)data to be classified |
Reimplemented from CMachine.
Definition at line 45 of file LinearStructuredOutputMachine.cpp.
|
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.
dict | dictionary of parameters to be built. |
Definition at line 1243 of file SGObject.cpp.
|
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.
Definition at line 1360 of file SGObject.cpp.
Locks the machine on given labels and data. After this call, only train_locked and apply_locked may be called
Only possible if supports_locking() returns true
labs | labels used for locking |
features | features used for locking |
Reimplemented in CKernelMachine.
Definition at line 120 of file Machine.cpp.
|
virtualinherited |
Unlocks a locked machine and restores previous state
Reimplemented in CKernelMachine.
Definition at line 151 of file Machine.cpp.
|
virtualinherited |
A deep copy. All the instance variables will also be copied.
Definition at line 200 of file SGObject.cpp.
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.
other | object to compare with |
accuracy | accuracy to use for comparison (optional) |
tolerant | allows linient check on float equality (within accuracy) |
Definition at line 1264 of file SGObject.cpp.
|
virtual |
get classifier type
Reimplemented from CMachine.
Definition at line 62 of file StochasticSOSVM.cpp.
int32_t get_debug_multiplier | ( | ) | const |
Definition at line 221 of file StochasticSOSVM.cpp.
|
inherited |
|
inherited |
|
inherited |
|
inherited |
|
inherited |
Definition at line 186 of file StructuredOutputMachine.cpp.
|
virtualinherited |
float64_t get_lambda | ( | ) | const |
Definition at line 201 of file StochasticSOSVM.cpp.
|
virtualinherited |
returns type of problem machine solves
Reimplemented in CNeuralNetwork, CRandomForest, CCHAIDTree, CCARTree, and CBaseMulticlassMachine.
|
inherited |
|
inherited |
get structured model
Definition at line 50 of file StructuredOutputMachine.cpp.
|
inherited |
Definition at line 1135 of file SGObject.cpp.
|
inherited |
Returns description of a given parameter string, if it exists. SG_ERROR otherwise
param_name | name of the parameter |
Definition at line 1159 of file SGObject.cpp.
|
inherited |
Returns index of model selection parameter with provided index
param_name | name of model selection parameter |
Definition at line 1172 of file SGObject.cpp.
|
virtual |
Reimplemented from CLinearStructuredOutputMachine.
Definition at line 52 of file StochasticSOSVM.h.
int32_t get_num_iter | ( | ) | const |
Definition at line 211 of file StochasticSOSVM.cpp.
uint32_t get_rand_seed | ( | ) | const |
Definition at line 231 of file StochasticSOSVM.cpp.
|
inherited |
|
inherited |
get surrogate loss function
Definition at line 91 of file StructuredOutputMachine.cpp.
|
inherited |
get verbose
Definition at line 203 of file StructuredOutputMachine.cpp.
|
inherited |
|
virtualinherited |
If the SGSerializable is a class template then TRUE will be returned and GENERIC is set to the type of the generic.
generic | set to the type of the generic if returning TRUE |
Definition at line 297 of file SGObject.cpp.
|
protectedvirtualinherited |
check whether the labels is valid.
Subclasses can override this to implement their check of label types.
lab | the labels being checked, guaranteed to be non-NULL |
Reimplemented in CNeuralNetwork, CCARTree, CCHAIDTree, CGaussianProcessRegression, and CBaseMulticlassMachine.
|
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)
file_version | parameter version of the file |
current_version | version from which mapping begins (you want to use Version::get_version_parameter() for this in most cases) |
file | file to load from |
prefix | prefix for members |
Definition at line 704 of file SGObject.cpp.
|
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
param_info | information of parameter |
file_version | parameter version of the file, must be <= provided parameter version |
file | file to load from |
prefix | prefix for members |
Definition at line 545 of file SGObject.cpp.
|
virtualinherited |
Load this object from file. If it will fail (returning FALSE) then this object will contain inconsistent data and should not be used!
file | where to load from |
prefix | prefix for members |
param_version | (optional) a parameter version different to (this is mainly for testing, better do not use) |
Definition at line 374 of file SGObject.cpp.
|
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.
ShogunException | will be thrown if an error occurs. |
Reimplemented in CKernel, CWeightedDegreePositionStringKernel, CList, CAlphabet, CLinearHMM, CGaussianKernel, CInverseMultiQuadricKernel, CCircularKernel, and CExponentialKernel.
Definition at line 1062 of file SGObject.cpp.
|
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.
ShogunException | will 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 1057 of file SGObject.cpp.
|
inherited |
problem type
|
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
param_base | set of TParameter instances that are mapped to the provided target parameter infos |
base_version | version of the parameter base |
target_param_infos | set of SGParamInfo instances that specify the target parameter base |
Definition at line 742 of file SGObject.cpp.
|
protectedvirtualinherited |
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
param_base | set of TParameter instances to use for migration |
target | parameter info for the resulting TParameter |
Definition at line 949 of file SGObject.cpp.
|
protectedvirtualinherited |
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)
param_base | set of TParameter instances to use for migration |
target | parameter info for the resulting TParameter |
replacement | (used as output) here the TParameter instance which is returned by migration is created into |
to_migrate | the only source that is used for migration |
old_name | with this parameter, a name change may be specified |
Definition at line 889 of file SGObject.cpp.
|
virtualinherited |
Definition at line 263 of file SGObject.cpp.
|
inherited |
prints all parameter registered for model selection and their type
Definition at line 1111 of file SGObject.cpp.
|
virtualinherited |
prints registered parameters out
prefix | prefix for members |
Definition at line 309 of file SGObject.cpp.
|
virtualinherited |
computes the value of the risk function and sub-gradient at given point
subgrad | Subgradient computed at given point W |
W | Given weight vector |
info | Helper info for multiple cutting plane models algorithm |
rtype | The type of structured risk |
Definition at line 157 of file StructuredOutputMachine.cpp.
|
protectedvirtualinherited |
1-slack formulation and margin rescaling
subgrad | Subgradient computed at given point W |
W | Given weight vector |
info | Helper info for multiple cutting plane models algorithm |
Definition at line 139 of file StructuredOutputMachine.cpp.
|
protectedvirtualinherited |
1-slack formulation and slack rescaling
subgrad | Subgradient computed at given point W |
W | Given weight vector |
info | Helper info for multiple cutting plane models algorithm |
Definition at line 145 of file StructuredOutputMachine.cpp.
|
protectedvirtualinherited |
customized risk type
subgrad | Subgradient computed at given point W |
W | Given weight vector |
info | Helper info for multiple cutting plane models algorithm |
Definition at line 151 of file StructuredOutputMachine.cpp.
|
protectedvirtualinherited |
n-slack formulation and margin rescaling
The value of the risk is evaluated as
\[ R({\bf w}) = \sum_{i=1}^{m} \max_{y \in \mathcal{Y}} \left[ \ell(y_i, y) + \langle {\bf w}, \Psi(x_i, y) - \Psi(x_i, y_i) \rangle \right] \]
The subgradient is by Danskin's theorem given as
\[ R'({\bf w}) = \sum_{i=1}^{m} \Psi(x_i, \hat{y}_i) - \Psi(x_i, y_i), \]
where \( \hat{y}_i \) is the most violated label, i.e.
\[ \hat{y}_i = \arg\max_{y \in \mathcal{Y}} \left[ \ell(y_i, y) + \langle {\bf w}, \Psi(x_i, y) \rangle \right] \]
subgrad | Subgradient computed at given point W |
W | Given weight vector |
info | Helper info for multiple cutting plane models algorithm |
Definition at line 97 of file StructuredOutputMachine.cpp.
|
protectedvirtualinherited |
n-slack formulation and slack rescaling
subgrad | Subgradient computed at given point W |
W | Given weight vector |
info | Helper info for multiple cutting plane models algorithm |
Definition at line 133 of file StructuredOutputMachine.cpp.
|
virtualinherited |
Save this object to file.
file | where to save the object; will be closed during returning if PREFIX is an empty string. |
prefix | prefix for members |
param_version | (optional) a parameter version different to (this is mainly for testing, better do not use) |
Definition at line 315 of file SGObject.cpp.
|
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.
ShogunException | will be thrown if an error occurs. |
Reimplemented in CKernel.
Definition at line 1072 of file SGObject.cpp.
|
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.
ShogunException | will 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 1067 of file SGObject.cpp.
void set_debug_multiplier | ( | int32_t | multiplier | ) |
set frequency of debug outputs
multiplier | debug multiplier |
Definition at line 226 of file StochasticSOSVM.cpp.
|
inherited |
|
inherited |
Definition at line 42 of file SGObject.cpp.
|
inherited |
Definition at line 47 of file SGObject.cpp.
|
inherited |
Definition at line 52 of file SGObject.cpp.
|
inherited |
Definition at line 57 of file SGObject.cpp.
|
inherited |
Definition at line 62 of file SGObject.cpp.
|
inherited |
Definition at line 67 of file SGObject.cpp.
|
inherited |
Definition at line 72 of file SGObject.cpp.
|
inherited |
Definition at line 77 of file SGObject.cpp.
|
inherited |
Definition at line 82 of file SGObject.cpp.
|
inherited |
Definition at line 87 of file SGObject.cpp.
|
inherited |
Definition at line 92 of file SGObject.cpp.
|
inherited |
Definition at line 97 of file SGObject.cpp.
|
inherited |
Definition at line 102 of file SGObject.cpp.
|
inherited |
Definition at line 107 of file SGObject.cpp.
|
inherited |
Definition at line 112 of file SGObject.cpp.
|
inherited |
set generic type to T
|
inherited |
|
inherited |
set the parallel object
parallel | parallel object to use |
Definition at line 243 of file SGObject.cpp.
|
inherited |
set the version object
version | version object to use |
Definition at line 284 of file SGObject.cpp.
|
virtualinherited |
set labels
lab | labels |
Reimplemented from CMachine.
Definition at line 67 of file StructuredOutputMachine.cpp.
void set_lambda | ( | float64_t | lbda | ) |
set regularization const
lbda | regularization const lambda |
Definition at line 206 of file StochasticSOSVM.cpp.
|
inherited |
set maximum training time
t | maximimum training time |
Definition at line 90 of file Machine.cpp.
|
inherited |
set structured model
model | structured model to set |
Definition at line 43 of file StructuredOutputMachine.cpp.
void set_num_iter | ( | int32_t | num_iter | ) |
set max number of iterations
num_iter | number of iterations |
Definition at line 216 of file StochasticSOSVM.cpp.
void set_rand_seed | ( | uint32_t | rand_seed | ) |
|
inherited |
|
virtualinherited |
Setter for store-model-features-after-training flag
store_model | whether model should be stored after training |
Definition at line 115 of file Machine.cpp.
|
inherited |
set surrogate loss function
loss | loss function to set |
Definition at line 84 of file StructuredOutputMachine.cpp.
|
inherited |
set verbose NOTE that track verbose information including primal objectives, training errors and duality gaps will make the training 2x or 3x slower.
verbose | flag enabling/disabling verbose information |
Definition at line 198 of file StructuredOutputMachine.cpp.
set w (useful for modular interfaces)
w | weight vector to set |
Definition at line 35 of file LinearStructuredOutputMachine.cpp.
|
virtualinherited |
A shallow copy. All the SGObject instance variables will be simply assigned and SG_REF-ed.
Reimplemented in CGaussianKernel.
Definition at line 194 of file SGObject.cpp.
|
virtualinherited |
Stores feature data of underlying model. Does nothing because Linear machines store the normal vector of the separating hyperplane and therefore the model anyway
Reimplemented from CMachine.
Definition at line 78 of file LinearStructuredOutputMachine.cpp.
|
virtualinherited |
Reimplemented in CKernelMachine, and CMultitaskLinearMachine.
|
virtualinherited |
train machine
data | training 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. |
Reimplemented in CRelaxedTree, CAutoencoder, CSGDQN, and COnlineSVMSGD.
Definition at line 47 of file Machine.cpp.
Trains a locked machine on a set of indices. Error if machine is not locked
NOT IMPLEMENTED
indices | index vector (of locked features) that is used for training |
Reimplemented in CKernelMachine, and CMultitaskLinearMachine.
|
protectedvirtual |
train primal SO-SVM
data | training data |
Reimplemented from CMachine.
Definition at line 67 of file StochasticSOSVM.cpp.
|
protectedvirtualinherited |
returns whether machine require labels for training
Reimplemented in COnlineLinearMachine, CHierarchical, CLinearLatentMachine, CVwConditionalProbabilityTree, CConditionalProbabilityTree, and CLibSVMOneClass.
|
inherited |
unset generic type
this has to be called in classes specializing a template class
Definition at line 304 of file SGObject.cpp.
|
virtualinherited |
Updates the hash of current parameter combination
Definition at line 250 of file SGObject.cpp.
|
inherited |
io
Definition at line 496 of file SGObject.h.
|
protectedinherited |
|
inherited |
parameters wrt which we can compute gradients
Definition at line 511 of file SGObject.h.
|
inherited |
Hash of parameter values
Definition at line 517 of file SGObject.h.
|
protectedinherited |
the helper that records primal objectives, duality gaps etc
Definition at line 223 of file StructuredOutputMachine.h.
|
protectedinherited |
|
protectedinherited |
the model that contains the application dependent modules
Definition at line 214 of file StructuredOutputMachine.h.
|
inherited |
model selection parameters
Definition at line 508 of file SGObject.h.
|
inherited |
map for different parameter versions
Definition at line 514 of file SGObject.h.
|
inherited |
parameters
Definition at line 505 of file SGObject.h.
|
protectedinherited |
|
protectedinherited |
|
protectedinherited |
the surrogate loss, for SOSVM, fixed to Hinge loss, other non-convex losses such as Ramp loss are also applicable, will be extended in the future
Definition at line 220 of file StructuredOutputMachine.h.
|
protectedinherited |
verbose outputs and statistics
Definition at line 226 of file StructuredOutputMachine.h.
weight vector
Definition at line 82 of file LinearStructuredOutputMachine.h.
|
inherited |
parallel
Definition at line 499 of file SGObject.h.
|
inherited |
version
Definition at line 502 of file SGObject.h.