evaluate_class_mlpT_evaluate_class_mlpEvaluateClassMlpEvaluateClassMlpevaluate_class_mlp (Operator)

Name

evaluate_class_mlpT_evaluate_class_mlpEvaluateClassMlpEvaluateClassMlpevaluate_class_mlp — Calculate the evaluation of a feature vector by a multilayer perceptron.

Signature

evaluate_class_mlp( : : MLPHandle, Features : Result)

Description

evaluate_class_mlpevaluate_class_mlpEvaluateClassMlpEvaluateClassMlpEvaluateClassMlpevaluate_class_mlp computes the result ResultResultResultResultresultresult of evaluating the feature vector FeaturesFeaturesFeaturesFeaturesfeaturesfeatures with the multilayer perceptron (MLP) MLPHandleMLPHandleMLPHandleMLPHandleMLPHandlemlphandle. The formulas used for the evaluation are described with create_class_mlpcreate_class_mlpCreateClassMlpCreateClassMlpCreateClassMlpcreate_class_mlp. Before calling evaluate_class_mlpevaluate_class_mlpEvaluateClassMlpEvaluateClassMlpEvaluateClassMlpevaluate_class_mlp, the MLP must be trained with train_class_mlptrain_class_mlpTrainClassMlpTrainClassMlpTrainClassMlptrain_class_mlp.

If the MLP is used for regression (function approximation), i.e., if (OutputFunction = 'linear'"linear""linear""linear""linear""linear"), ResultResultResultResultresultresult is the value of the function at the coordinate FeaturesFeaturesFeaturesFeaturesfeaturesfeatures. For OutputFunction = 'logistic'"logistic""logistic""logistic""logistic""logistic" and 'softmax'"softmax""softmax""softmax""softmax""softmax", the values in ResultResultResultResultresultresult can be interpreted as probabilities. Hence, for OutputFunction = 'logistic'"logistic""logistic""logistic""logistic""logistic" the elements of ResultResultResultResultresultresult represent the probabilities of the presence of the respective independent attributes. Typically, a threshold of 0.5 is used to decide whether the attribute is present or not. Depending on the application, other thresholds may be used as well. For OutputFunction = 'softmax'"softmax""softmax""softmax""softmax""softmax" usually the position of the maximum value of ResultResultResultResultresultresult is interpreted as the class of the feature vector, and the corresponding value as the probability of the class. In this case, classify_class_mlpclassify_class_mlpClassifyClassMlpClassifyClassMlpClassifyClassMlpclassify_class_mlp should be used instead of evaluate_class_mlpevaluate_class_mlpEvaluateClassMlpEvaluateClassMlpEvaluateClassMlpevaluate_class_mlp because classify_class_mlpclassify_class_mlpClassifyClassMlpClassifyClassMlpClassifyClassMlpclassify_class_mlp directly returns the class and corresponding probability.

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

MLPHandleMLPHandleMLPHandleMLPHandleMLPHandlemlphandle (input_control) class_mlp → (handle)

MLP handle.

FeaturesFeaturesFeaturesFeaturesfeaturesfeatures (input_control) real-array → (real)

Feature vector.

ResultResultResultResultresultresult (output_control) real-array → (real)

Result of evaluating the feature vector with the MLP.

Result

If the parameters are valid, the operator evaluate_class_mlpevaluate_class_mlpEvaluateClassMlpEvaluateClassMlpEvaluateClassMlpevaluate_class_mlp returns the value TRUE. If necessary, an exception is raised.

Possible Predecessors

train_class_mlptrain_class_mlpTrainClassMlpTrainClassMlpTrainClassMlptrain_class_mlp, read_class_mlpread_class_mlpReadClassMlpReadClassMlpReadClassMlpread_class_mlp

Alternatives

classify_class_mlpclassify_class_mlpClassifyClassMlpClassifyClassMlpClassifyClassMlpclassify_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