C++ API Reference for Intel® Data Analytics Acceleration Library 2019 Update 5

References | Namespaces | Classes | Enumerations
Quality Metrics for Binary Classification Algorithms

Contains classes for computing the binary confusion matrix. More...

References

 Batch
 

Namespaces

 daal::algorithms::classifier::quality_metric::binary_confusion_matrix
 Contains classes for computing the binary confusion matrix.
 
 daal::algorithms::classifier::quality_metric::binary_confusion_matrix::interface1
 Contains version 1.0 of the Intel(R) Data Analytics Acceleration Library (Intel(R) DAAL) interface.
 

Classes

struct  Parameter
 Parameters for the binary confusion matrix compute() method. More...
 
class  Input
 Base class for input objects of the binary confusion matrix algorithm. More...
 
class  Result
 Results obtained with the compute() method of the binary confusion matrix algorithm in the batch processing mode. More...
 

Enumerations

enum  Method { defaultDense = 0 }
 
enum  InputId { predictedLabels, groundTruthLabels }
 
enum  ResultId { confusionMatrix, binaryMetrics }
 
enum  BinaryMetricsId {
  accuracy, precision, recall, fscore,
  specificity, AUC
}
 

Enumeration Type Documentation

enum BinaryMetricsId

Available values stored in a numeric table corresponding to the ResultId::binaryMatrix index

Enumerator
accuracy 

Accuracy

precision 

Precision

recall 

Recall

fscore 

F-score

specificity 

Specificity

AUC 

Area under the curve (AUC). Ability to avoid false classification

enum InputId

Available identifiers of input objects for the binary confusion matrix algorithm

Enumerator
predictedLabels 

Labels computed in the prediction stage of the classification algorithm

groundTruthLabels 

Expected labels

enum Method

Available methods for computing the binary confusion matrix

Enumerator
defaultDense 

Default method

enum ResultId

Available identifiers of results of the binary confusion matrix algorithm

Enumerator
confusionMatrix 

Binary confusion matrix

binaryMetrics 

Table that contains quality metrics (that is, precision, recall, etc.) for binary classifiers

For more complete information about compiler optimizations, see our Optimization Notice.