Units

Classifier	Contains classification components.
DBClassifier	Database interface for classification.

Components

TDBClassifier	Use the component to perform classification on the data stored in any TDataset descendant.
TKNearestNeighbors	Implementation of the K-NN (K Nearest Neighbors) classification algorithm.
TLinearClassifier	Implementation of the "Linear classifier" classifier.
TNaiveBayes	Implementation of the "Naive Bayes" classification algorithm.

Classes

TAttributeDataType	Defines the mapping of field types to attribute types.
TBClassesList	Holds a list of classes for Naive Bayes classification algorithm.
TBClassesListItem	Classification class definition for the Naive Bayes classification algorithm.
TClassesList	The abstract object holds a list of known classification classes.
TClassesListItem	Abstract class for items of a list of clasification classes.
TClassifier	Abstract class for encapsulation of different classification algorithms.
TDBClassifier	Use the component to perform classification on the data stored in any TDataset descendant.
TFieldColumn	Connects the classification attribute to a dataset field.
TFieldColumns	Used to store classsification attribute definitions.
TKNearestNeighbors	Implementation of the K-NN (K Nearest Neighbors) classification algorithm.
TLClassesList	Holds a list of classes for the Linear Classifier classification algorithm.
TLClassesListItem	Classification class for the "Linear classifier" classification algorithm.
TLinearClassifier	Implementation of the "Linear classifier" classifier.
TNaiveBayes	Implementation of the "Naive Bayes" classification algorithm.
TStatisticClassifier	Abstract classifier class for statistical classifiers.

Types

PAPointer
PKRecord	Pointer to the TKRecord type.
TAttributeKind	Defines the type of the attribute.
TAttributeType	Defines types of classification attributes.
TDiscreteRecord	Record type used to store parameteres learned for discrete attributes.
TDiscreteRecordArray
TDistanceModel	Distance model used by the K-NN algorithm.
TExampleRecord	Record type used by K-NN (K Nearest Neighbors) classification algorithm to store the entire learn database in memory.
TFloatRecord	Record type used to store parameteres learned for continuous attributes.
TFloatRecordArray
TIndexNextRecord	Event type used by TClassifier.
TIndexRecord	Used by the TClassifier to estimate the quality of the attributes Used by the TClassifier to estimate the quality of the attributes.
TKFloatSpan	Record type used by K-NN (K Nearest Neighbors) classification algorithm to store the information about the range of real valued attributes.
TKNNParams	Record type used by K-NN (K Nearest Neighbors) classification algorithm to control the use of attributes.
TKRecord	Record type used by K-NN (K Nearest Neighbors) classification algorithm to store the nearest neighbors.
TNameNextRecord	Event type used by TClassifier.
TOnClassifyDataset	Event type triggered before the classification is run.
TOnClassifyTest	Event type used to determine classification accuracy.
TOnLearnDataset	Event type triggered before the learning is run.
TPopBookmark	Event type used to get bookmark position.
TPushBookmark	Event type used to save bookmark position.

Routines

Mean	Compute average of the a array.
StdDev	Compute standard deviation.