STK::SemiLinearAAModel Class Reference [Project AAModels (Data Analysis with auto-associative]

Base class for all Semi-Linear AutoAssociative models (SLAAM). More...

#include <STK_SemiLinearAAModel.h>

Inherits STK::IAAModel.

Inherited by STK::LinearAAModel, and STK::PcaModel.

List of all members.

Public Member Functions
virtual	~SemiLinearAAModel ()
const Index &	index () const
Protected Member Functions
	SemiLinearAAModel (Index *p_index)
virtual void	projection ()
virtual void	wprojection ()
Protected Attributes
Index *	p_index_
IRegress< Matrix, Matrix > *	p_regress_

---

Detailed Description

Base class for all Semi-Linear AutoAssociative models (SLAAM).

A SLAAM is a (probabilistic) model of the form

in .

is an input d dimensional random variable and is the residual random variable. The (generalized) principal components are real random variables estimated by a projection of the centered data on a subspace

where the form an orthonormal set of vector.

The regression functions are specifics to the model.

The principal Component Analysis (PCA) is the most common AA models.

The pure virtual method to be implemented in the derived class are

   // regression without weights
   virtual void regression();
   // weighted regression
   virtual void wregression();
   // prediction of the data set
   virtual void prediction();

The data set is set to the Index outside this class as an Index can be shared by various IAAModel.

Definition at line 87 of file STK_SemiLinearAAModel.h.

---

Constructor & Destructor Documentation

STK::SemiLinearAAModel::SemiLinearAAModel

(

Index *

p_index

)

[protected]

Constructor. The containers workData_, p_predicted_, p_residuals_ and mean_ are initialized to theirs dimensions.

The data set to process have been set to p_index outside the class.

Parameters:

p_index

the index to use in order to obtain the axis

Definition at line 45 of file STK_SemiLinearAAModel.cpp.

                                    : IAAModel(p_index->p_data())

STK::SemiLinearAAModel::~SemiLinearAAModel

(

)

[virtual]

virtual destuctor.

Definition at line 52 of file STK_SemiLinearAAModel.cpp.

---

Member Function Documentation

const Index& STK::SemiLinearAAModel::index

(

)

const [inline]

Get the index used in the computation of the axis

Returns:

a reference on the Index object

Definition at line 123 of file STK_SemiLinearAAModel.h.

References p_index_.

{ return *p_index_;}

void STK::SemiLinearAAModel::projection

(

)

[protected, virtual]

compute the projection

Implements STK::IAAModel.

Definition at line 56 of file STK_SemiLinearAAModel.cpp.

{ }

/* compute the projection **/
void SemiLinearAAModel::projection()
{
  // compute axis
  p_index_->run(dim_);
  // remove existing  container
  if (p_projData_) delete p_projData_;
  // resize result
  p_projData_ = new Matrix(workData_.rangeVe(), p_index_->axis().rangeHo());

void STK::SemiLinearAAModel::wprojection

(

)

[protected, virtual]

compute the weighted projection

Implements STK::IAAModel.

Definition at line 69 of file STK_SemiLinearAAModel.cpp.

{
  // compute axis
  p_index_->run(p_weights_, dim_);
  // remove existing  container
  if (p_projData_) delete p_projData_;
  // resize result
  p_projData_ = new Matrix(workData_.rangeVe(), p_index_->axis().rangeHo());

---

Member Data Documentation

Index* STK::SemiLinearAAModel::p_index_ [protected]

the input Index used in order to obtain the axis.

Definition at line 91 of file STK_SemiLinearAAModel.h.

Referenced by index().

IRegress<Matrix, Matrix>* STK::SemiLinearAAModel::p_regress_ [protected]

the Regression method used in order to adjust the model.

Definition at line 93 of file STK_SemiLinearAAModel.h.

---

The documentation for this class was generated from the following files:

• STK_SemiLinearAAModel.h
• STK_SemiLinearAAModel.cpp