PLearn 0.1
Public Member Functions | Static Public Member Functions | Public Attributes | Static Public Attributes | Static Protected Member Functions | Private Types | Private Member Functions | Private Attributes
PLearn::AutoLinearRegressor Class Reference

The first sentence should be a BRIEF DESCRIPTION of what the class does. More...

#include <AutoLinearRegressor.h>

Inheritance diagram for PLearn::AutoLinearRegressor:
Inheritance graph
[legend]
Collaboration diagram for PLearn::AutoLinearRegressor:
Collaboration graph
[legend]

List of all members.

Public Member Functions

 AutoLinearRegressor ()
 Default constructor.
virtual int outputsize () const
 Returns the size of this learner's output, (which typically may depend on its inputsize(), targetsize() and set options).
virtual void forget ()
 (Re-)initializes the PLearner in its fresh state (that state may depend on the 'seed' option) and sets 'stage' back to 0 (this is the stage of a fresh learner!).
virtual void train ()
 The role of the train method is to bring the learner up to stage==nstages, updating the train_stats collector with training costs measured on-line in the process.
virtual void computeOutput (const Vec &input, Vec &output) const
 Computes the output from the input.
virtual void computeCostsFromOutputs (const Vec &input, const Vec &output, const Vec &target, Vec &costs) const
 Computes the costs from already computed output.
virtual TVec< std::string > getTestCostNames () const
 Returns the names of the costs computed by computeCostsFromOutpus (and thus the test method).
virtual TVec< std::string > getTrainCostNames () const
 Returns the names of the objective costs that the train method computes and for which it updates the VecStatsCollector train_stats.
virtual string classname () const
virtual OptionListgetOptionList () const
virtual OptionMapgetOptionMap () const
virtual RemoteMethodMapgetRemoteMethodMap () const
virtual AutoLinearRegressordeepCopy (CopiesMap &copies) const
virtual void build ()
 Finish building the object; just call inherited::build followed by build_()
virtual void makeDeepCopyFromShallowCopy (CopiesMap &copies)
 Transforms a shallow copy into a deep copy.

Static Public Member Functions

static string _classname_ ()
static OptionList_getOptionList_ ()
static RemoteMethodMap_getRemoteMethodMap_ ()
static Object_new_instance_for_typemap_ ()
static bool _isa_ (const Object *o)
static void _static_initialize_ ()
static const PPathdeclaringFile ()

Public Attributes

bool include_bias
 ### declare public option fields (such as build options) here Start your comments with Doxygen-compatible comments such as //! Whether to include a bias term in the regression (true by default)
real min_weight_decay
real weight_decay
Mat weights
Vec mean_target

Static Public Attributes

static StaticInitializer _static_initializer_

Static Protected Member Functions

static void declareOptions (OptionList &ol)
 Declares the class options.

Private Types

typedef PLearner inherited

Private Member Functions

void build_ ()
 This does the actual building.

Private Attributes

real weights_norm
 Sum of squares of weights.
Vec extendedinput

Detailed Description

The first sentence should be a BRIEF DESCRIPTION of what the class does.

Place the rest of the class programmer documentation here. Doxygen supports Javadoc-style comments. See http://www.doxygen.org/manual.html

Todo:
Write class to-do's here if there are any.
Deprecated:
Write deprecated stuff here if there is any. Indicate what else should be used instead.

Definition at line 57 of file AutoLinearRegressor.h.


Member Typedef Documentation

Reimplemented from PLearn::PLearner.

Definition at line 59 of file AutoLinearRegressor.h.


Constructor & Destructor Documentation

PLearn::AutoLinearRegressor::AutoLinearRegressor ( )

Default constructor.

Definition at line 54 of file AutoLinearRegressor.cc.

    : include_bias(false),
      min_weight_decay(1e-6),
      weight_decay(0.0)
{
}

Member Function Documentation

string PLearn::AutoLinearRegressor::_classname_ ( ) [static]

Reimplemented from PLearn::PLearner.

Definition at line 52 of file AutoLinearRegressor.cc.

OptionList & PLearn::AutoLinearRegressor::_getOptionList_ ( ) [static]

Reimplemented from PLearn::PLearner.

Definition at line 52 of file AutoLinearRegressor.cc.

RemoteMethodMap & PLearn::AutoLinearRegressor::_getRemoteMethodMap_ ( ) [static]

Reimplemented from PLearn::PLearner.

Definition at line 52 of file AutoLinearRegressor.cc.

bool PLearn::AutoLinearRegressor::_isa_ ( const Object o) [static]

Reimplemented from PLearn::PLearner.

Definition at line 52 of file AutoLinearRegressor.cc.

Object * PLearn::AutoLinearRegressor::_new_instance_for_typemap_ ( ) [static]

Reimplemented from PLearn::Object.

Definition at line 52 of file AutoLinearRegressor.cc.

StaticInitializer AutoLinearRegressor::_static_initializer_ & PLearn::AutoLinearRegressor::_static_initialize_ ( ) [static]

Reimplemented from PLearn::PLearner.

Definition at line 52 of file AutoLinearRegressor.cc.

void PLearn::AutoLinearRegressor::build ( ) [virtual]

Finish building the object; just call inherited::build followed by build_()

Reimplemented from PLearn::PLearner.

Definition at line 121 of file AutoLinearRegressor.cc.

References PLearn::PLearner::build(), and build_().

Here is the call graph for this function:

void PLearn::AutoLinearRegressor::build_ ( ) [private]

This does the actual building.

Reimplemented from PLearn::PLearner.

Definition at line 106 of file AutoLinearRegressor.cc.

Referenced by build().

{
    // ### This method should do the real building of the object,
    // ### according to set 'options', in *any* situation.
    // ### Typical situations include:
    // ###  - Initial building of an object from a few user-specified options
    // ###  - Building of a "reloaded" object: i.e. from the complete set of
    // ###    all serialised options.
    // ###  - Updating or "re-building" of an object after a few "tuning"
    // ###    options have been modified.
    // ### You should assume that the parent class' build_() has already been
    // ### called.
}

Here is the caller graph for this function:

string PLearn::AutoLinearRegressor::classname ( ) const [virtual]

Reimplemented from PLearn::Object.

Definition at line 52 of file AutoLinearRegressor.cc.

void PLearn::AutoLinearRegressor::computeCostsFromOutputs ( const Vec input,
const Vec output,
const Vec target,
Vec costs 
) const [virtual]

Computes the costs from already computed output.

Implements PLearn::PLearner.

Definition at line 248 of file AutoLinearRegressor.cc.

References PLearn::powdistance(), PLearn::TVec< T >::resize(), weight_decay, and weights_norm.

{
    // Compute the costs from *already* computed output. 
    costs.resize(2);
    real squared_loss = powdistance(output,target);
    costs[0] = squared_loss + weight_decay*weights_norm;
    costs[1] = squared_loss;
}

Here is the call graph for this function:

void PLearn::AutoLinearRegressor::computeOutput ( const Vec input,
Vec output 
) const [virtual]

Computes the output from the input.

Reimplemented from PLearn::PLearner.

Definition at line 230 of file AutoLinearRegressor.cc.

References extendedinput, include_bias, PLearn::PLearner::inputsize(), mean_target, outputsize(), PLearn::product(), PLearn::TVec< T >::resize(), PLearn::TVec< T >::subVec(), and weights.

{
    // Compute the output from the input
    int nout = outputsize();
    output.resize(nout);
    if(!include_bias)        
        product(output,weights,input);
    else
    {   
        int nin = inputsize();
        extendedinput.resize(1+nin);
        extendedinput.subVec(1,nin) << input;
        extendedinput[0] = 1.0;
        product(output,weights,extendedinput);
    }
    output += mean_target;
}

Here is the call graph for this function:

void PLearn::AutoLinearRegressor::declareOptions ( OptionList ol) [static, protected]

Declares the class options.

Reimplemented from PLearn::PLearner.

Definition at line 61 of file AutoLinearRegressor.cc.

References PLearn::OptionBase::buildoption, PLearn::declareOption(), PLearn::PLearner::declareOptions(), include_bias, PLearn::OptionBase::learntoption, mean_target, min_weight_decay, weight_decay, and weights.

{
    // ### Declare all of this object's options here.
    // ### For the "flags" of each option, you should typically specify
    // ### one of OptionBase::buildoption, OptionBase::learntoption or
    // ### OptionBase::tuningoption. If you don't provide one of these three,
    // ### this option will be ignored when loading values from a script.
    // ### You can also combine flags, for example with OptionBase::nosave:
    // ### (OptionBase::buildoption | OptionBase::nosave)

    // ### ex:
    // declareOption(ol, "myoption", &AutoLinearRegressor::myoption,
    //               OptionBase::buildoption,
    //               "Help text describing this option");
    // ...

    declareOption(ol, "include_bias", &AutoLinearRegressor::include_bias,
                  OptionBase::buildoption,
                  "Whether to include a bias term in the regression \n"
                  "Note: this is currently ignored.\n");

    declareOption(ol, "min_weight_decay", &AutoLinearRegressor::min_weight_decay,
                  OptionBase::buildoption, 
                  "The minimum weight decay to try.");


    declareOption(ol, "weight_decay", &AutoLinearRegressor::weight_decay,
                  OptionBase::learntoption, 
                  "The weight decay is the factor that multiplies the \n"
                  "squared norm of the parameters in the loss function.\n"
                  "It is automatically tuned by the algorithm \n");

    declareOption(ol, "weights", &AutoLinearRegressor::weights,
                  OptionBase::learntoption, 
                  "The weight matrix, which are the parameters computed by "
                  "training the regressor.\n");

    declareOption(ol, "mean_target", &AutoLinearRegressor::mean_target,
                  OptionBase::learntoption,
                  "The mean of the target (used as a default bias).");

    // Now call the parent class' declareOptions
    inherited::declareOptions(ol);
}

Here is the call graph for this function:

static const PPath& PLearn::AutoLinearRegressor::declaringFile ( ) [inline, static]

Reimplemented from PLearn::PLearner.

Definition at line 145 of file AutoLinearRegressor.h.

:
    //#####  Protected Options  ###############################################
AutoLinearRegressor * PLearn::AutoLinearRegressor::deepCopy ( CopiesMap copies) const [virtual]

Reimplemented from PLearn::PLearner.

Definition at line 52 of file AutoLinearRegressor.cc.

void PLearn::AutoLinearRegressor::forget ( ) [virtual]

(Re-)initializes the PLearner in its fresh state (that state may depend on the 'seed' option) and sets 'stage' back to 0 (this is the stage of a fresh learner!).

Reimplemented from PLearn::PLearner.

Definition at line 152 of file AutoLinearRegressor.cc.

References PLearn::PLearner::forget(), PLearn::TMat< T >::resize(), PLearn::PLearner::stage, and weights.

{
    weights.resize(0,0);
    stage = 0;
    inherited::forget();
}

Here is the call graph for this function:

OptionList & PLearn::AutoLinearRegressor::getOptionList ( ) const [virtual]

Reimplemented from PLearn::Object.

Definition at line 52 of file AutoLinearRegressor.cc.

OptionMap & PLearn::AutoLinearRegressor::getOptionMap ( ) const [virtual]

Reimplemented from PLearn::Object.

Definition at line 52 of file AutoLinearRegressor.cc.

RemoteMethodMap & PLearn::AutoLinearRegressor::getRemoteMethodMap ( ) const [virtual]

Reimplemented from PLearn::Object.

Definition at line 52 of file AutoLinearRegressor.cc.

TVec< string > PLearn::AutoLinearRegressor::getTestCostNames ( ) const [virtual]

Returns the names of the costs computed by computeCostsFromOutpus (and thus the test method).

Implements PLearn::PLearner.

Definition at line 258 of file AutoLinearRegressor.cc.

References PLearn::TVec< T >::push_back().

{
    TVec<string> names;
    names.push_back("mse+penalty");
    names.push_back("mse");
    return names;
}

Here is the call graph for this function:

TVec< string > PLearn::AutoLinearRegressor::getTrainCostNames ( ) const [virtual]

Returns the names of the objective costs that the train method computes and for which it updates the VecStatsCollector train_stats.

Implements PLearn::PLearner.

Definition at line 266 of file AutoLinearRegressor.cc.

References PLearn::TVec< T >::push_back().

{
    TVec<string> names;
    names.push_back("GCV_mse");
    names.push_back("mse");
    return names;
}

Here is the call graph for this function:

void PLearn::AutoLinearRegressor::makeDeepCopyFromShallowCopy ( CopiesMap copies) [virtual]

Transforms a shallow copy into a deep copy.

Reimplemented from PLearn::PLearner.

Definition at line 128 of file AutoLinearRegressor.cc.

References PLearn::deepCopyField(), extendedinput, PLearn::PLearner::makeDeepCopyFromShallowCopy(), mean_target, and weights.

{
    // ### Call deepCopyField on all "pointer-like" fields
    // ### that you wish to be deepCopied rather than
    // ### shallow-copied.
    // ### ex:
    // deepCopyField(trainvec, copies);

    // ### Remove this line when you have fully implemented this method.
    //PLERROR("AutoLinearRegressor::makeDeepCopyFromShallowCopy not fully (correctly) implemented yet!");

    inherited::makeDeepCopyFromShallowCopy(copies);
    deepCopyField(weights, copies);
    deepCopyField(mean_target, copies);
    deepCopyField(extendedinput, copies);

}

Here is the call graph for this function:

int PLearn::AutoLinearRegressor::outputsize ( ) const [virtual]

Returns the size of this learner's output, (which typically may depend on its inputsize(), targetsize() and set options).

Implements PLearn::PLearner.

Definition at line 147 of file AutoLinearRegressor.cc.

References PLearn::PLearner::targetsize().

Referenced by computeOutput().

{
    return targetsize();
}

Here is the call graph for this function:

Here is the caller graph for this function:

void PLearn::AutoLinearRegressor::train ( ) [virtual]

The role of the train method is to bring the learner up to stage==nstages, updating the train_stats collector with training costs measured on-line in the process.

Implements PLearn::PLearner.

Definition at line 159 of file AutoLinearRegressor.cc.

References PLearn::TMat< T >::column(), PLearn::columnMean(), PLearn::dot(), PLearn::TVec< T >::fill(), PLearn::hconcat(), i, include_bias, PLearn::PLearner::initTrain(), PLearn::TMat< T >::length(), mean_target, min_weight_decay, PLearn::multiplyAcc(), PLearn::TMat< T >::resize(), PLearn::TVec< T >::resize(), PLearn::PLearner::stage, PLearn::TMat< T >::subMatColumns(), PLearn::sum(), PLearn::TMat< T >::toVecCopy(), PLearn::PLearner::train_set, PLearn::PLearner::train_stats, weight_decay, PLearn::weightedRidgeRegressionByGCV(), weights, and weights_norm.

{
    // The role of the train method is to bring the learner up to
    // stage==nstages, updating train_stats with training costs measured
    // on-line in the process.

    // This generic PLearner method does a number of standard stuff useful for
    // (almost) any learner, and return 'false' if no training should take
    // place. See PLearner.h for more details.
    if (!initTrain())
        return;

    if(stage<1)
    {
        // clear statistics of previous epoch
        train_stats->forget();
        
        int ninputs = train_set->inputsize();
        int ntargets = train_set->targetsize();
        int nweights = train_set->weightsize();

        Mat tset = train_set->toMatCopy();
        int l = tset.length();
        
        Mat X = tset.subMatColumns(0,ninputs);
        Mat Y = tset.subMatColumns(ninputs, ntargets);
        Vec gamma; // the weights

        if (include_bias)
        {
            Mat col_ones = Mat(l, 1, 1.0);
            X = hconcat(col_ones, X);
        }

        mean_target.resize(ntargets);
        mean_target.fill(0);

        if(nweights!=0)
        {
            gamma = tset.column(ninputs+ntargets).toVecCopy();
            for(int i=0;  i<l; i++)
                multiplyAcc(mean_target, Y(i), gamma[i]);
            mean_target /= sum(gamma);
        }
        else
            columnMean(Y, mean_target);
        Y -= mean_target;

        int insize = ninputs + (include_bias ? 1 : 0); 
        //weights.resize(insize, ntargets);
        weights.resize(ntargets, insize);
        real best_GCV;
      
        weight_decay = weightedRidgeRegressionByGCV(X, Y, gamma, weights, best_GCV, min_weight_decay);

        //Mat weights_excluding_biases = weights.subMatRows(include_bias? 1 : 0, ninputs);
        Mat weights_excluding_biases = weights.subMatColumns(include_bias? 1 : 0, ninputs);
        weights_norm = dot(weights_excluding_biases,weights_excluding_biases);

        //Vec trcosts(1);
        Vec trcosts(2);
        trcosts[0] = best_GCV;
        trcosts[1] = best_GCV;
        train_stats->update(trcosts);

        ++stage;
        train_stats->finalize(); // finalize statistics for this epoch
    }
}

Here is the call graph for this function:


Member Data Documentation

Reimplemented from PLearn::PLearner.

Definition at line 145 of file AutoLinearRegressor.h.

Definition at line 181 of file AutoLinearRegressor.h.

Referenced by computeOutput(), and makeDeepCopyFromShallowCopy().

### declare public option fields (such as build options) here Start your comments with Doxygen-compatible comments such as //! Whether to include a bias term in the regression (true by default)

Definition at line 67 of file AutoLinearRegressor.h.

Referenced by computeOutput(), declareOptions(), and train().

Definition at line 69 of file AutoLinearRegressor.h.

Referenced by declareOptions(), and train().

Definition at line 71 of file AutoLinearRegressor.h.

Referenced by computeCostsFromOutputs(), declareOptions(), and train().

Sum of squares of weights.

Definition at line 180 of file AutoLinearRegressor.h.

Referenced by computeCostsFromOutputs(), and train().


The documentation for this class was generated from the following files:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines