select_feature_set_knnT_select_feature_set_knnSelectFeatureSetKnnSelectFeatureSetKnnselect_feature_set_knn (算子名称)

名称

select_feature_set_knnT_select_feature_set_knnSelectFeatureSetKnnSelectFeatureSetKnnselect_feature_set_knn — Selects an optimal subset from a set of features to solve a certain classification problem.

参数签名

描述

select_feature_set_knnselect_feature_set_knnSelectFeatureSetKnnSelectFeatureSetKnnSelectFeatureSetKnnselect_feature_set_knn selects an optimal subset from a set of features to solve a certain classification problem. The classification problem has to be specified with annotated training data in ClassTrainDataHandleClassTrainDataHandleClassTrainDataHandleClassTrainDataHandleclassTrainDataHandleclass_train_data_handle and will be classified by a a K-最近邻 classifier. Details of the properties of this classifier can be found in create_class_knncreate_class_knnCreateClassKnnCreateClassKnnCreateClassKnncreate_class_knn.

The result of 该算子 is a trained classifier that is returned in KNNHandleKNNHandleKNNHandleKNNHandleKNNHandleknnhandle. Additionally, the list of indices or names of the selected features is returned in SelectedFeatureIndicesSelectedFeatureIndicesSelectedFeatureIndicesSelectedFeatureIndicesselectedFeatureIndicesselected_feature_indices. To use this classifier, calculate for new input data all features mentioned in SelectedFeatureIndicesSelectedFeatureIndicesSelectedFeatureIndicesSelectedFeatureIndicesselectedFeatureIndicesselected_feature_indices and pass them to the classifier.

A possible application of this operator can be a comparison of different parameter sets for certain feature extraction techniques. Another application is to search for a property that is discriminating between different classes of parts or classes of errors.

To define the features that should be selected from ClassTrainDataHandleClassTrainDataHandleClassTrainDataHandleClassTrainDataHandleclassTrainDataHandleclass_train_data_handle, the dimensions of the feature vectors in ClassTrainDataHandleClassTrainDataHandleClassTrainDataHandleClassTrainDataHandleclassTrainDataHandleclass_train_data_handle can be grouped into subfeatures by calling set_feature_lengths_class_train_dataset_feature_lengths_class_train_dataSetFeatureLengthsClassTrainDataSetFeatureLengthsClassTrainDataSetFeatureLengthsClassTrainDataset_feature_lengths_class_train_data. A subfeature can contain several subsequent elements of a feature vector. 该算子 decides for each of these subfeatures, if it is better to use it for the classification or leave it out.

The indices of the selected subfeatures are returned in SelectedFeatureIndicesSelectedFeatureIndicesSelectedFeatureIndicesSelectedFeatureIndicesselectedFeatureIndicesselected_feature_indices. If names were set in set_feature_lengths_class_train_dataset_feature_lengths_class_train_dataSetFeatureLengthsClassTrainDataSetFeatureLengthsClassTrainDataSetFeatureLengthsClassTrainDataset_feature_lengths_class_train_data, these names are returned instead of the indices. If set_feature_lengths_class_train_dataset_feature_lengths_class_train_dataSetFeatureLengthsClassTrainDataSetFeatureLengthsClassTrainDataSetFeatureLengthsClassTrainDataset_feature_lengths_class_train_data was not called for ClassTrainDataHandleClassTrainDataHandleClassTrainDataHandleClassTrainDataHandleclassTrainDataHandleclass_train_data_handle before, each element of the feature vector is considered as a subfeature.

The selection method SelectionMethodSelectionMethodSelectionMethodSelectionMethodselectionMethodselection_method is either a greedy search 'greedy'"greedy""greedy""greedy""greedy""greedy" (iteratively add the feature with highest gain) or the dynamically oscillating search 'greedy_oscillating'"greedy_oscillating""greedy_oscillating""greedy_oscillating""greedy_oscillating""greedy_oscillating" (add the feature with highest gain and test then if any of the already added features can be left out without great loss). The method 'greedy'"greedy""greedy""greedy""greedy""greedy" is generally preferable, since it is faster. Only in cases when the subfeatures are low-dimensional or redundant, the method 'greedy_oscillating'"greedy_oscillating""greedy_oscillating""greedy_oscillating""greedy_oscillating""greedy_oscillating" should be chosen.

The optimization criterion is the classification rate of a two-fold cross-validation of the training data. The best achieved value is returned in ScoreScoreScoreScorescorescore.

The k-NN classifier can be parameterized using the following values in GenParamNameGenParamNameGenParamNameGenParamNamegenParamNamegen_param_name and GenParamValueGenParamValueGenParamValueGenParamValuegenParamValuegen_param_value:

'num_neighbors'"num_neighbors""num_neighbors""num_neighbors""num_neighbors""num_neighbors":

The number of minimally evaluated nodes, increase this value for high dimensional data.

Possible values: '1'"1""1""1""1""1", '2'"2""2""2""2""2", '5'"5""5""5""5""5", '10'"10""10""10""10""10"

Default value: '1'"1""1""1""1""1"

'num_trees'"num_trees""num_trees""num_trees""num_trees""num_trees":

Number of search trees in the k-NN classifier

Possible values: '1'"1""1""1""1""1", '4'"4""4""4""4""4", '10'"10""10""10""10""10"

Default value: '4'"4""4""4""4""4"

注意

This operator may take considerable time, depending on the size of the data set in the training file, and the number of features.

Please note, that this operator should not be called, if only a small set of training data is available. Due to the risk of overfitting the operator select_feature_set_knnselect_feature_set_knnSelectFeatureSetKnnSelectFeatureSetKnnSelectFeatureSetKnnselect_feature_set_knn may deliver a classifier with a very high score. However, the classifier may perform poorly when tested.

运行信息

• 多线程类型:可重入(与非独占操作符并行运行)。
• 多线程作用域:全局(可以从任何线程调用)。
• Automatically parallelized on internal data level.

This operator returns a handle. Note that the state of an instance of this handle type may be changed by specific operators even though the handle is used as an input parameter by those operators.

参数表

例程 (HDevelop)

* Find out which of the two features distinguishes two Classes
NameFeature1 := 'Good Feature'
NameFeature2 := 'Bad Feature'
LengthFeature1 := 3
LengthFeature2 := 2
* Create training data
create_class_train_data (LengthFeature1+LengthFeature2,\
  ClassTrainDataHandle)
* Define the features which are in the training data
set_feature_lengths_class_train_data (ClassTrainDataHandle, [LengthFeature1,\
  LengthFeature2], [NameFeature1, NameFeature2])
* Add training data
*                                                         |Feat1| |Feat2|
add_sample_class_train_data (ClassTrainDataHandle, 'row', [1,1,1,  2,1  ], 0)
add_sample_class_train_data (ClassTrainDataHandle, 'row', [2,2,2,  2,1  ], 1)
add_sample_class_train_data (ClassTrainDataHandle, 'row', [1,1,1,  3,4  ], 0)
add_sample_class_train_data (ClassTrainDataHandle, 'row', [2,2,2,  3,4  ], 1)
add_sample_class_train_data (ClassTrainDataHandle, 'row', [0,0,1,  5,6  ], 0)
add_sample_class_train_data (ClassTrainDataHandle, 'row', [2,3,2,  5,6  ], 1)
* Add more data
* ...
* Select the better feature with the k-NN classifier
select_feature_set_knn (ClassTrainDataHandle, 'greedy', [], [], KNNHandle,\
  SelectedFeatureKNN, Score)
* Use the classifier
* ...

结果

如果参数均有效，算子 select_feature_set_knnselect_feature_set_knnSelectFeatureSetKnnSelectFeatureSetKnnSelectFeatureSetKnnselect_feature_set_knn 返回值 2 ( H_MSG_TRUE) . 如有必要，将引发异常。

可能的前置算子

create_class_train_datacreate_class_train_dataCreateClassTrainDataCreateClassTrainDataCreateClassTrainDatacreate_class_train_data, add_sample_class_train_dataadd_sample_class_train_dataAddSampleClassTrainDataAddSampleClassTrainDataAddSampleClassTrainDataadd_sample_class_train_data, set_feature_lengths_class_train_dataset_feature_lengths_class_train_dataSetFeatureLengthsClassTrainDataSetFeatureLengthsClassTrainDataSetFeatureLengthsClassTrainDataset_feature_lengths_class_train_data

可能的后置算子

classify_class_knnclassify_class_knnClassifyClassKnnClassifyClassKnnClassifyClassKnnclassify_class_knn

可替代算子

select_feature_set_mlpselect_feature_set_mlpSelectFeatureSetMlpSelectFeatureSetMlpSelectFeatureSetMlpselect_feature_set_mlp, select_feature_set_svmselect_feature_set_svmSelectFeatureSetSvmSelectFeatureSetSvmSelectFeatureSetSvmselect_feature_set_svm, select_feature_set_gmmselect_feature_set_gmmSelectFeatureSetGmmSelectFeatureSetGmmSelectFeatureSetGmmselect_feature_set_gmm

参考其它

select_feature_set_trainf_knnselect_feature_set_trainf_knnSelectFeatureSetTrainfKnnSelectFeatureSetTrainfKnnSelectFeatureSetTrainfKnnselect_feature_set_trainf_knn, gray_featuresgray_featuresGrayFeaturesGrayFeaturesGrayFeaturesgray_features, region_featuresregion_featuresRegionFeaturesRegionFeaturesRegionFeaturesregion_features

模块

Foundation

---