Classes | |
class | ApplyConcepts |
This class handles the logic of applying trained models to a dataset. More... | |
class | ApplyConceptsHelper |
class | ApplyConceptsHelperFeatures |
class | ApplyConceptsHelperFik |
class | ApplyConceptsHelperKernels |
class | AreaUnderRocCurve |
This class can be applied to a ScoreTable to compute the area under the ROC curve. More... | |
class | AveragePrecision |
This class can be applied to a ScoreTable to compute the average precision. More... | |
class | BalancedAveragePrecision |
class | Classifier |
This is a baseclass for classifiers. More... | |
class | ClassifierEvaluator |
This class works together with ParameterSearcher (through baseclass ParameterEvaluator). More... | |
class | Evaluation |
Baseclass for evaluation of ranked lists. More... | |
class | Factory |
This class' purpose is to generalise some of the setup that was shared by some applications. More... | |
class | FikSvm |
class | KernelMatrix |
First setup of a KernelMatrix class, work in progress. More... | |
class | ParameterEvaluator |
Abstract base class for classes that the ParameterSearcher can work with. More... | |
class | ParameterSearcher |
This class searches a range of parameters for the best setting. More... | |
class | PrecisionAtN |
This class can be applied to a ScoreTable to compute the precision at a given N. More... | |
class | PrecomputeTask |
PrecomputeTask: manages information about what to compute. More... | |
class | RecallAtN |
This class can be applied to a ScoreTable to compute the recall at a given N. More... | |
class | Svm |
SVM is a well known Classifier, see internet for details. More... | |
class | SvmProblemBuilder |
This class does the actual conversion from feature and annotation data to the data format used by SvmLib. More... | |
class | TestBestFile |
This class tests some feature functionality. More... | |
class | TestSvm |
class | TrainDataSrc |
Abstraction of training data. More... | |
class | TrainDataSrcFeature |
class | TrainDataSrcFeatureTable |
class | TrainDataSrcKernelDistributed |
this class assumes some other cpus serve the distributed matrix. More... | |
class | TrainDataSrcKernelMatrix |
Wraps a kernel matrix (of type Matrix::Mat*). More... | |
Functions | |
ILOG_VAR_INIT (ApplyConcepts, Impala.Core.Training) | |
ILOG_VAR_INIT (ApplyConceptsHelperFeatures, Impala.Core.Training) | |
ILOG_VAR_INIT (ApplyConceptsHelperFik, Impala.Core.Training) | |
ILOG_VAR_INIT (ApplyConceptsHelperKernels, Impala.Core.Training) | |
ILOG_VAR_INIT (AveragePrecision, Impala.Core.Training) | |
int | BestFileDiff (Util::PropertySet *set1, Util::PropertySet *set2) |
int | BestFileReferenceDiff (Util::PropertySet *set1, Util::PropertySet *set2) |
ILOG_VAR_INIT (ClassifierEvaluator, Impala.Core.Training) | |
template<class FType, class PrecomputeTaskT> | |
void | ComputeKernelMatrix (PrecomputeTaskT *pt, int slabWidth, bool GPU) |
Kernel matrix computation. | |
bool | Equals (const svm_parameter *p1, const svm_parameter *p2) |
All Equals functions compare the objects (they do not test for 0 pointers). | |
bool | EqualsForModel (const svm_parameter *p1, const svm_parameter *p2) |
bool | Equals (const svm_node *sv1, const svm_node *sv2) |
bool | Equals (const svm_problem *p1, const svm_problem *p2) |
bool | Equals (const svm_model *m1, const svm_model *m2) |
ILOG_VAR_INIT (Factory, Impala.Core.Training) | |
ILOG_VAR_INIT (FikSvm, Impala.Core.Training) | |
ILOG_VAR_INIT (KernelMatrix, Impala.Core.Training) | |
Matrix::DistributedAccess * | LoadDistributedAccess (String model, String kernel, Core::Database::RawDataSet *set, Core::Database::RawDataSet *set2, int startNode, int nodeCount) |
Evaluation * | MakeEvaluation (const std::string &description, Table::AnnotationTable *annotation) |
Factory function for Evaluation objects. | |
ILOG_VAR_INIT (ParameterSearcher, Impala.Core.Training) | |
ILOG_VAR_INIT (PrecomputeTask, Impala.Core.Training) | |
ILOG_VAR_INIT (Svm, Impala.Core.Training) | |
bool | Equals (const Svm *svm1, const Svm *svm2) |
int | Diff (const Svm *svm1, const Svm *svm2) |
ILOG_VAR_INIT (SvmProblemBuilder, Impala.Core.Training) | |
ILOG_CLASS_INIT (TestBestFile, Impala.Core.Training) | |
CPPUNIT_TEST_SUITE_REGISTRATION (TestBestFile) | |
CPPUNIT_TEST_SUITE_REGISTRATION (TestSvm) | |
ILOG_VAR_INIT (TrainDataSrc, Impala.Core.Training) | |
ILOG_VAR_INIT (TrainDataSrcFeature, Impala.Core.Training) | |
ILOG_VAR_INIT (TrainDataSrcFeatureTable, Impala.Core.Training) | |
ILOG_VAR_INIT (TrainDataSrcKernelDistributed, Impala.Core.Training) | |
Variables | |
const double | cPrecision = 0.000001 |
since svm lib saves in ascii with 6 decimal numbers, we can't expect higher precision than this |