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

References | Namespaces | Classes | Enumerations
Objective Function

References

 Batch
 
 Cross-entropy loss Algorithm
 Contains classes for computing the Cross-entropy loss objective function.
 
 Logistic loss Algorithm
 Contains classes for computing the Logistic loss objective function.
 
 Mean Squared Error Algorithm
 Contains classes for computing the Mean squared error objective function.
 
 Objective function with precomputed characteristics
 Contains classes for the Objective function with precomputed characteristics.
 
 Sum of Functions
 Contains classes for computing the Sum of functions.
 

Namespaces

 daal::algorithms::optimization_solver::objective_function
 Contains classes for computing the Objective function.
 
 daal::algorithms::optimization_solver::objective_function::interface1
 Contains version 1.0 of the Intel(R) Data Analytics Acceleration Library (Intel(R) DAAL) interface.
 

Classes

struct  Parameter
 Parameter for the Objective function More...
 
class  Input
 Input objects for the Objective function More...
 
class  Result
 Provides methods to access final results obtained with the compute() method of the Objective function in the batch processing mode. More...
 

Enumerations

enum  InputId { argument }
 
enum  ResultToComputeId {
  gradient = 0x00000001ULL, value = 0x00000002ULL, hessian = 0x00000004ULL, nonSmoothTermValue = 0x00000008ULL,
  proximalProjection = 0x00000010ULL, lipschitzConstant = 0x00000020ULL, componentOfGradient = 0x00000040ULL, componentOfHessianDiagonal = 0x00000080ULL,
  componentOfProximalProjection = 0x00000100ULL
}
 
enum  ResultId {
  gradientIdx, valueIdx, hessianIdx, nonSmoothTermValueIdx,
  proximalProjectionIdx, lipschitzConstantIdx, componentOfGradientIdx, componentOfHessianDiagonalIdx,
  componentOfProximalProjectionIdx
}
 

Enumeration Type Documentation

enum InputId

Available identifiers of input objects of the Objective function

Enumerator
argument 

Numeric table of size 1 x p with input argument of the objective function

enum ResultId

Available identifiers of results of the Objective function

Enumerator
gradientIdx 

Index of the gradient numeric table in the result collection

valueIdx 

Index of the value numeric table in the result collection

hessianIdx 

Index of the hessian numeric table in the result collection

nonSmoothTermValueIdx 

Index of the nonSmoothTermValue numeric table in the result collection

proximalProjectionIdx 

Index of the proximalProjection numeric table in the result collection

lipschitzConstantIdx 

Index of the lipschitzConstant numeric table in the result collection

componentOfGradientIdx 

Index of the componentOfGradient numeric table in the result collection

componentOfHessianDiagonalIdx 

Index of the componentOfHessianDiagonal numeric table in the result collection

componentOfProximalProjectionIdx 

Index of the componentOfProximalProjection numeric table in the result collection

enum ResultToComputeId

Available identifiers to specify the Objective function result

Enumerator
gradient 

Numeric table of size 1 x p with the gradient of the objective function in the given argument

value 

Numeric table of size 1 x 1 with the value of the objective function in the given argument

hessian 

Numeric table of size p x p with the hessian of the objective function in the given argument

nonSmoothTermValue 

Numeric table of size 1 x 1 with the value of non-smooth part of the objective function in the given argument

proximalProjection 

Numeric table of size 1 x p with the proximal projection of non-smooth part of the objective function in the given argument

lipschitzConstant 

Numeric table of size 1 x 1 with the Lipschitz constant of the objective function in the given argument

componentOfGradient 

Numeric table of size 1 x nDependentVariable with the gradient over certain feature of the objective function in the given argument

componentOfHessianDiagonal 

Numeric table of size 1 x nDependentVariable with the dioganal element of hession matrix over certain feature of the objective function in the given argument

componentOfProximalProjection 

Numeric table of size p x nDependentVariable with proximal projection of certain of the objective function in the given argument

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