C++ API Reference for Intel® Data Analytics Acceleration Library 2020 Update 1

Namespaces | Enumerations
daal::algorithms::implicit_als::training Namespace Reference

Contains classes of the implicit ALS training algorithm.

Namespaces

 init
 Contains classes for the implicit ALS initialization algorithm.
 
 interface1
 Contains version 1.0 of the Intel(R) Data Analytics Acceleration Library (Intel(R) DAAL) interface.
 

Enumerations

enum  Method { defaultDense = 0, fastCSR = 1 }
 
enum  NumericTableInputId { data }
 
enum  ModelInputId { inputModel = lastNumericTableInputId + 1 }
 
enum  PartialModelInputId { partialModel }
 
enum  MasterInputId { inputOfStep2FromStep1 }
 
enum  DistributedPartialResultStep1Id { outputOfStep1ForStep2 }
 
enum  DistributedPartialResultStep2Id { outputOfStep2ForStep4 }
 
enum  Step3LocalCollectionInputId { partialModelBlocksToNode = lastDistributedPartialResultStep2Id + 1, inputOfStep3FromInit = partialModelBlocksToNode }
 
enum  Step3LocalNumericTableInputId { offset = lastStep3LocalCollectionInputId + 1 }
 
enum  DistributedPartialResultStep3Id { outputOfStep3ForStep4 }
 
enum  Step4LocalPartialModelsInputId { partialModels }
 
enum  Step4LocalNumericTableInputId { partialData = lastStep4LocalPartialModelsInputId + 1, inputOfStep4FromStep2 }
 
enum  DistributedPartialResultStep4Id { outputOfStep4ForStep1, outputOfStep4ForStep3 = outputOfStep4ForStep1, outputOfStep4 = outputOfStep4ForStep3 }
 
enum  ResultId { model }
 Available identifiers of the results of the implicit ALS training algorithm. More...
 

Enumeration Type Documentation

Available types of partial results of the implicit ALS training algorithm in the first step of the distributed processing mode

Enumerator
outputOfStep1ForStep2 

Partial results of the implicit ALS training algorithm computed in the first step and to be transferred to the second step of the distributed processing mode

Available types of partial results of the implicit ALS training algorithm in the second step of the distributed processing mode

Enumerator
outputOfStep2ForStep4 

Partial results of the implicit ALS training algorithm computed in the second step and to be transferred to the fourth step of the distributed processing mode

Available types of partial results of the implicit ALS training algorithm in the third step of the distributed processing mode

Enumerator
outputOfStep3ForStep4 

Partial results of the implicit ALS training algorithm computed in the third step and to be transferred to the fourth step of the distributed processing mode

Available types of partial results of the implicit ALS training algorithm in the fourth step of the distributed processing mode

Enumerator
outputOfStep4ForStep1 

Partial results of the implicit ALS training algorithm computed in the fourth step and to be transferred to the first step of the distributed processing mode

outputOfStep4ForStep3 

Partial results of the implicit ALS training algorithm computed in the fourth step and to be transferred to the third step of the distributed processing mode

outputOfStep4 

Partial results of the implicit ALS training algorithm computed in the fourth step and to be used in implicit ALS PartialModel-based prediction

Partial results obtained in the previous step and required by the second step of the distributed processing mode

Enumerator
inputOfStep2FromStep1 

Partial results of the implicit ALS training algorithm computed in the first step and to be transferred to the second step of the distributed processing mode

enum Method

Available methods for training the implicit ALS model

Enumerator
defaultDense 

Default: method proposed by Hu, Koren, Volinsky for input data stored in the dense format

fastCSR 

Method proposed by Hu, Koren, Volinsky for input data stored in the compressed sparse row (CSR) format

Available identifiers of input model objects for the implicit ALS training algorithm

Enumerator
inputModel 

Initial model that contains initialized factors

Available identifiers of input numeric table objects for the implicit ALS training algorithm

Enumerator
data 

Input data table that contains ratings

Available identifiers of input partial model objects of the implicit ALS training algorithm

Enumerator
partialModel 

Partial model that contains factors obtained in the previous step of the distributed processing mode

enum ResultId

Enumerator
model 

Implicit ALS model

Available identifiers of input data collection objects for the implicit ALS training algorithm in the third step of the distributed processing mode

Enumerator
partialModelBlocksToNode 
Deprecated:
This item will be removed in a future release. Use inputOfStep3FromInit instead.

Partial results of the implicit ALS initialization algorithm to be transferred to the third step of the implicit ALS training algorithm in the distributed processing mode

inputOfStep3FromInit 

Partial results of the implicit ALS initialization algorithm to be transferred to the third step of the implicit ALS training algorithm in the distributed processing mode

Available identifiers of input numeric table objects for the implicit ALS training algorithm in the third step of the distributed processing mode

Enumerator
offset 

Pointer to the 1x1 numeric table that holds the global index of the starting row of the input partial model

Available identifiers of input numeric table objects for the implicit ALS training algorithm in the fourth step of the distributed processing mode

Enumerator
partialData 

Pointer to the CSR numeric table that holds a block of either users or items from the input data set

inputOfStep4FromStep2 

Pointer to the nFactors x nFactors numeric table computed in the second step of the distributed processing mode

Available identifiers of input key-value data collection objects for the implicit ALS training algorithm in the fourth step of the distributed processing mode

Enumerator
partialModels 

Key-value data collection that contains partial models consisting of user factors/item factors computed in the third step of the distributed processing mode. Each element of the collection contains an object of the PartialModel class.

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