Back to Adaptive Vision Library website

You are here: Start » Function Reference » Support Vector Machines » SVM_ClassifyMultiple



Classifies input points based on trained model


void avl::SVM_ClassifyMultiple
	const avl::SvmModel& inSvmModel,
	const atl::Array<atl::Array<float> >& inVectorArray,
	atl::Array< int >& outPredictions,
	atl::Optional<atl::Array< int >& > outModelClasses = atl::NIL,
	atl::Optional<atl::Array<atl::Array<float>>& > outClassProbabilities = atl::NIL


Name Type Default Description
inSvmModel const SvmModel& Input trained model
inVectorArray const Array<Array<float> >& Data vector array of unknown classes
outPredictions Arrayint >& Predicted classes
outModelClasses Optional<Arrayint >& > NIL All known model classes in order
outClassProbabilities Optional<Array<Array<float>>& > NIL For each data vector the probability of belonging to each class

Optional Outputs

The computation of following outputs can be switched off by passing value atl::NIL to these parameters: outModelClasses, outClassProbabilities.

Read more about Optional Outputs.


The operation predicts classes for the given data points. It takes an array of data vectors (inVectorArray) as an argument. Each vector has to be of the same size as vectors used for training the model. The operation outputs predicted class (outPredictions) for each data vector.
outModelClasses are all class labels encountered during training.
outClassProbabilities provides, for each vector, estimated probability of this vector belonging to each class. Precisely, in each array the value under index i denotes probability of given vector belonging to the class outModelClasses[i].


List of possible exceptions:

Error type Description
DomainError Empty vector array in SVM_ClassifyMultiple.
DomainError Incorrect vector size in SVM_Classify
DomainError Incorrect, uninitialized or not trained SvmModel in SVM_ClassifyMultiple.

See Also