classify_class_mlpT_classify_class_mlpClassifyClassMlpClassifyClassMlpclassify_class_mlp (Operator)

Name

classify_class_mlpT_classify_class_mlpClassifyClassMlpClassifyClassMlpclassify_class_mlp — Calculate the class of a feature vector by a multilayer perceptron.

Signature

classify_class_mlp( : : MLPHandle, Features, Num : Class, Confidence)

Description

classify_class_mlpclassify_class_mlpClassifyClassMlpClassifyClassMlpclassify_class_mlp computes the best NumNumNumnumnum classes of the feature vector FeaturesFeaturesFeaturesfeaturesfeatures with the multilayer perceptron (MLP) MLPHandleMLPHandleMLPHandleMLPHandlemlphandle and returns the classes in ClassClassClassclassValclass and the corresponding confidences (probabilities) of the classes in ConfidenceConfidenceConfidenceconfidenceconfidence. Before calling classify_class_mlpclassify_class_mlpClassifyClassMlpClassifyClassMlpclassify_class_mlp, the MLP must be trained with train_class_mlptrain_class_mlpTrainClassMlpTrainClassMlptrain_class_mlp.

classify_class_mlpclassify_class_mlpClassifyClassMlpClassifyClassMlpclassify_class_mlp can only be called if the MLP is used as a classifier with OutputFunctionOutputFunctionOutputFunctionoutputFunctionoutput_function = 'softmax'"softmax""softmax""softmax""softmax" (see create_class_mlpcreate_class_mlpCreateClassMlpCreateClassMlpcreate_class_mlp). Otherwise, an error message is returned. classify_class_mlpclassify_class_mlpClassifyClassMlpClassifyClassMlpclassify_class_mlp corresponds to a call to evaluate_class_mlpevaluate_class_mlpEvaluateClassMlpEvaluateClassMlpevaluate_class_mlp and an additional step that extracts the best NumNumNumnumnum classes. As described with evaluate_class_mlpevaluate_class_mlpEvaluateClassMlpEvaluateClassMlpevaluate_class_mlp, the output values of the MLP can be interpreted as probabilities of the occurrence of the respective classes. In most cases it should be sufficient to use NumNumNumnumnum = 1 in order to decide whether the probability of the best class is high enough. In some applications it may be interesting to also take the second best class into account (NumNumNumnumnum = 2), particularly if it can be expected that the classes show a significant degree of overlap.

Execution Information

Multithreading type: reentrant (runs in parallel with non-exclusive operators).
Multithreading scope: global (may be called from any thread).
Processed without parallelization.

Parameters

MLPHandleMLPHandleMLPHandleMLPHandlemlphandle (input_control) class_mlp → (handle)

MLP handle.

FeaturesFeaturesFeaturesfeaturesfeatures (input_control) real-array → (real)

Feature vector.

NumNumNumnumnum (input_control) integer-array → (integer)

Number of best classes to determine.

Default: 1

Suggested values: 1, 2, 3, 4, 5

ClassClassClassclassValclass (output_control) integer(-array) → (integer)

Result of classifying the feature vector with the MLP.

ConfidenceConfidenceConfidenceconfidenceconfidence (output_control) real(-array) → (real)

Confidence(s) of the class(es) of the feature vector.

Result

If the parameters are valid, the operator classify_class_mlpclassify_class_mlpClassifyClassMlpClassifyClassMlpclassify_class_mlp returns the value 2 ( H_MSG_TRUE) . If necessary, an exception is raised.

Possible Predecessors

train_class_mlptrain_class_mlpTrainClassMlpTrainClassMlptrain_class_mlp, read_class_mlpread_class_mlpReadClassMlpReadClassMlpread_class_mlp

Alternatives

apply_dl_classifierapply_dl_classifierApplyDlClassifierApplyDlClassifierapply_dl_classifier, evaluate_class_mlpevaluate_class_mlpEvaluateClassMlpEvaluateClassMlpevaluate_class_mlp

References

Christopher M. Bishop: “Neural Networks for Pattern Recognition”; Oxford University Press, Oxford; 1995.
Andrew Webb: “Statistical Pattern Recognition”; Arnold, London; 1999.

Module

Foundation

Operators