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

References | Namespaces | Classes | Enumerations

Contains classes of the K-Means algorithm. More...

References

 Batch
 
 Distributed
 

Namespaces

 daal::algorithms::kmeans
 Contains classes of the K-Means algorithm.
 
 daal::algorithms::kmeans::interface1
 Contains version 1.0 of the Intel(R) Data Analytics Acceleration Library (Intel(R) DAAL) interface.
 

Classes

struct  Parameter
 Parameters for the K-Means algorithm. More...
 
class  InputIface
 Interface for input objects for the the K-Means algorithm in the batch and distributed processing modes. More...
 
class  Input
 Input objects for the K-Means algorithm More...
 
class  PartialResult
 Partial results obtained with the compute() method of the K-Means algorithm in the batch processing mode. More...
 
class  Result
 Results obtained with the compute() method of the K-Means algorithm in the batch processing mode. More...
 
class  DistributedStep2MasterInput
 Input objects for the K-Means algorithm in the distributed processing mode More...
 

Enumerations

enum  Method { lloydDense = 0, defaultDense = 0, lloydCSR = 1 }
 
enum  DistanceType { euclidean }
 
enum  InputId { data, inputCentroids }
 Available identifiers of input objects for the K-Means algorithm. More...
 
enum  MasterInputId { partialResults }
 Available identifiers of input objects for the K-Means algorithm in the distributed processing mode. More...
 
enum  PartialResultId {
  nObservations, partialSums, partialObjectiveFunction, partialGoalFunction = partialObjectiveFunction,
  partialAssignments, partialCandidatesDistances, partialCandidatesCentroids
}
 Available identifiers of partial results of the K-Means algorithm in the distributed processing mode. More...
 
enum  ResultId {
  centroids, assignments, objectiveFunction, goalFunction = objectiveFunction,
  nIterations
}
 Available identifiers of results of the K-Means algorithm. More...
 

Enumeration Type Documentation

enum DistanceType

Supported distance types

Enumerator
euclidean 

Euclidean distance

enum InputId

Enumerator
data 

Input data table

inputCentroids 

Initial centroids for the algorithm

enum MasterInputId

Enumerator
partialResults 

Collection of partial results computed on local nodes

enum Method

Available methods of the K-Means algorithm

Enumerator
lloydDense 

Default: performance-oriented method, synonym of defaultDense

defaultDense 

Default: performance-oriented method, synonym of lloydDense

lloydCSR 

Implementation of the Lloyd algorithm for CSR numeric tables

enum PartialResultId

Enumerator
nObservations 

Table containing the number of observations assigned to centroids

partialSums 

Table containing the sum of observations assigned to centroids

partialObjectiveFunction 

Table containing an objective function value

partialGoalFunction 

Table containing an objective function value

Deprecated:
This item will be removed in a future release.
partialAssignments 

Table containing assignments of observations to particular clusters

partialCandidatesDistances 

Table containing goal function of observations most distant from their assigned cluster center

partialCandidatesCentroids 

Table containing observations most distant from their assigned cluster center

enum ResultId

Enumerator
centroids 

Table containing cluster centroids

assignments 

Table containing assignments of observations to particular clusters

objectiveFunction 

Table containing an objective function value

goalFunction 

Table containing an objective function value

Deprecated:
This item will be removed in a future release.
nIterations 

Table containing the number of executed iterations

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