classifier_multinom_naivebayes.hpp

Go to the documentation of this file.

#ifndef FSTCLASSIFIERMULTINOMNAIVEBAYES_H
#define FSTCLASSIFIERMULTINOMNAIVEBAYES_H
 /* 
=========================================================================
Copyright and license:
  * FST4 software (with the exception of any bundled or externally linked
    third-party components) is Copyright (C) 2000-present Institute of
    Information Theory and Automation of the Czech Academy of Sciences (UTIA).
  * FST4 is free software: you can redistribute it and/or modify it under
    the terms of the GNU Affero General Public License as published by the
    Free Software Foundation, version 3. The full text of the license is in
    the file LICENSE in the distribution root and at
    https://www.gnu.org/licenses/agpl-3.0.html.
  * FST4 is distributed in the hope that it will be useful, but WITHOUT ANY
    WARRANTY; without even the implied warranty of MERCHANTABILITY or
    FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public
    License for more details.
  * Closed-source use without the AGPL's copyleft obligations requires a
    commercial license from UTIA; in case of interest write to
    utia@utia.cas.cz. LICENSING.md in the distribution root explains both
    options, the contributor license agreement that every contribution
    requires, and the contacts (technical questions and pull requests:
    somol@utia.cas.cz, with "FST" in the subject line).
========================================================================== */
 
#include <memory>
#include <cassert>
#include <iostream>
#include <sstream>
#include <cmath>
#include <cstring> // memcpy
#include "error.hpp"
#include "global.hpp"
#include "classifier.hpp"
#include "model_multinom.hpp"
 
/*============== Template parameter type naming conventions ==============
--------- Numeric types: -------------------------------------------------
DATATYPE - data sample values - usually real numbers (but may be integers
          in text processing etc.)
REALTYPE - must be real numbers - for representing intermediate results of 
          calculations like mean, covariance etc.
IDXTYPE - index values for enumeration of data samples - (nonnegative) integers, 
          extent depends on numbers of samples in data
DIMTYPE - index values for enumeration of features (dimensions), or classes (not 
          class sizes) - (nonnegative) integers, usually lower extent than IDXTYPE, 
          but be aware of expressions like _classes*_features*_features ! 
          in linearized representations of feature matrices for all classes
BINTYPE - feature selection marker type - represents ca. <10 different feature 
          states (selected, deselected, sel./desel. temporarily 1st nested loop, 2nd...)
RETURNTYPE - criterion value: real value, but may be extended in future to support 
          multiple values 
--------- Class types: ---------------------------------------------------
SUBSET       - class of class type Subset 
CLASSIFIER   - class implementing interface defined in abstract class Classifier 
EVALUATOR    - class implementing interface defined in abstract class Sequential_Step 
DISTANCE     - class implementing interface defined in abstract class Distance 
DATAACCESSOR - class implementing interface defined in abstract class Data_Accessor 
INTERVALCONTAINER - class of class type Data_Intervaller 
CONTAINER    - STL container of class type Data_Interval  
========================================================================== */
 
namespace FST {

template<class RETURNTYPE, typename DATATYPE, typename REALTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>
class Classifier_Multinomial_NaiveBayes : public Clonable_Impl<Classifier_Multinomial_NaiveBayes<RETURNTYPE,DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR>,Classifier<RETURNTYPE,DIMTYPE,SUBSET,DATAACCESSOR> > {
public:
        typedef Clonable_Impl<Classifier_Multinomial_NaiveBayes,Classifier<RETURNTYPE,DIMTYPE,SUBSET,DATAACCESSOR> > parent_impl; 
        typedef std::shared_ptr<DATAACCESSOR> PDataAccessor; 
        typedef std::shared_ptr<SUBSET> const PSubset; 
        typedef typename DATAACCESSOR::PPattern PPattern; 
        Classifier_Multinomial_NaiveBayes();
        virtual ~Classifier_Multinomial_NaiveBayes() {notify("Classifier_Multinomial_NaiveBayes destructor.");}

        void enable_prelearn_mode(const PDataAccessor da);
        void disable_prelearn_mode() {_prelearn_mode=false;} 
        bool get_prelearn_mode() const {return _prelearn_mode;} 

        void initialize(const PDataAccessor da);
 
        virtual bool classify(DIMTYPE &cls, const PPattern &pattern) override;  // classifies pattern, returns the respective class index
        virtual bool classify_scores(std::vector<RETURNTYPE> &scores, const PPattern &pattern) override
        {       // per-class log-posteriors for ranking metrics (AUC); single features are unusable
                // (theta=1 in each class, see classify()) -> equal scores = uninformative ranking (AUC 0.5)
                if(_subfeatures==1) {scores.assign(_classes,(RETURNTYPE)0.0); return true;}
                DIMTYPE cls;
                if(!classify(cls,pattern)) return false;
                scores.resize(_classes);
                for(DIMTYPE c=0;c<_classes;c++) scores[c]=(RETURNTYPE)_Pcd[c];
                return true;
        }
        virtual bool test_confusion(std::vector<RETURNTYPE> &confusion, DIMTYPE &classes, const PDataAccessor da) override
        {       // mirrors test()'s "single features unusable" convention (correct=0, count=1):
                // one off-diagonal count -> every confusion metric scores 0, like accuracy does
                if(_subfeatures==1) {
                        classes=da->getNoOfClasses();
                        if(classes<2) return false;
                        confusion.assign((size_t)classes*(size_t)classes,(RETURNTYPE)0.0);
                        confusion[1]=1.0; // true class 0 predicted as class 1
                        return true;
                }
                return Classifier<RETURNTYPE,DIMTYPE,SUBSET,DATAACCESSOR>::test_confusion(confusion,classes,da);
        }
        virtual bool train(const PDataAccessor da, const PSubset sub) override; // learns from designated training part of data
        virtual bool test(RETURNTYPE &result, const PDataAccessor da) override; // estimates accuracy using designated test data
        
        DIMTYPE get_n() const {assert(_model); return _model->get_n();} 
        DIMTYPE get_d() const {assert(_model); return _model->get_d();} 
 
 
        
        virtual std::ostream& print(std::ostream& os) const override {os << "Classifier_Multinomial_NaiveBayes()"; return os;}
private:
        friend parent_impl; 
        Classifier_Multinomial_NaiveBayes(const Classifier_Multinomial_NaiveBayes& cmnb); 
protected:
        std::unique_ptr<Model_Multinomial<DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR> > _model; 
 
        DIMTYPE _classes; 
        DIMTYPE _features; 
        std::unique_ptr<REALTYPE[]> _Pcd; 
 
private:
        bool _prelearn_mode; 
        std::unique_ptr<DIMTYPE[]> _index; 
        DIMTYPE _subfeatures; 
};
 
template<class RETURNTYPE, typename DATATYPE, typename REALTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>
Classifier_Multinomial_NaiveBayes<RETURNTYPE,DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR>::Classifier_Multinomial_NaiveBayes(const Classifier_Multinomial_NaiveBayes& cmnb) :
        _classes(cmnb._classes),
        _features(cmnb._features),
        _prelearn_mode(cmnb._prelearn_mode),
        _subfeatures(cmnb._subfeatures)
{
        notify("Classifier_Multinomial_NaiveBayes copy-constructor.");
        if(cmnb._model) _model.reset(new Model_Multinomial<DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR>(*cmnb._model));
        if(_classes>0)
        {
                _Pcd.reset(new REALTYPE[_classes]); memcpy((void *)_Pcd.get(),(void *)(cmnb._Pcd).get(),sizeof(REALTYPE)*_classes);
        }
        if(_features>0) // guards against memcpy from cmnb._index==null when copying a never-initialize()d instance (_features>0 implies _index was allocated -- see initialize())
        {
                _index.reset(new DIMTYPE[_features]); memcpy((void *)_index.get(),(void *)(cmnb._index).get(),sizeof(DIMTYPE)*_features);
        }
}
 
 
 
template<class RETURNTYPE, typename DATATYPE, typename REALTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>
Classifier_Multinomial_NaiveBayes<RETURNTYPE,DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR>::Classifier_Multinomial_NaiveBayes()
{
        notify("Classifier_Multinomial_NaiveBayes constructor.");
        _model.reset(new Model_Multinomial<DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR>);
        _classes=0;
        _features=0;
        _subfeatures=0;
        _prelearn_mode=false;
}
 
template<class RETURNTYPE, typename DATATYPE, typename REALTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>
void Classifier_Multinomial_NaiveBayes<RETURNTYPE,DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR>::enable_prelearn_mode(const PDataAccessor da)
{
        notify("Classifier_Multinomial_NaiveBayes::init_prelearn_mode().");
        assert(_model);
        assert(da);
        assert(da->getNoOfFeatures()>0);
        assert(da->getNoOfClasses()>0);
        _model->learn(da);
        _prelearn_mode=true;
}
 
template<class RETURNTYPE, typename DATATYPE, typename REALTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>
void Classifier_Multinomial_NaiveBayes<RETURNTYPE,DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR>::initialize(const PDataAccessor da)
{
        notify("Classifier_Multinomial_NaiveBayes::initialize().");
        assert(da);
        assert(da->getNoOfFeatures()>0);
        assert(da->getNoOfClasses()>0);
        
        if(!_Pcd || _classes!=da->getNoOfClasses()) _Pcd.reset(new REALTYPE[da->getNoOfClasses()]);
        _classes=da->getNoOfClasses();
 
        if(!_index || _features!=da->getNoOfFeatures()) _index.reset(new DIMTYPE[da->getNoOfFeatures()]);
        _features=da->getNoOfFeatures();
}
 
template<class RETURNTYPE, typename DATATYPE, typename REALTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>
bool Classifier_Multinomial_NaiveBayes<RETURNTYPE,DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR>::train(const PDataAccessor da, const PSubset sub)
{
        // NOTE: initialize() is called unconditionally on every train(), so the buffers always match
        //       the accessor currently in use even when the wrapper switches datasets or splits. It
        //       only (re)allocates when the class or feature count changed, so the repeated call is
        //       cheap; it could be skipped while evaluating several subsets over one unchanged split.
        initialize(da);
 
        notify("Classifier_Multinomial_NaiveBayes::train().");
        assert(_model);
        assert(da);
        if(_prelearn_mode) assert(get_n()>0);
        assert(da->getNoOfFeatures()>0);
        assert(sub);
        assert(sub->get_frozen_mode()==false);
        if(_prelearn_mode) assert(sub->get_n_raw()==get_n());
        assert(sub->get_n_raw()==da->getNoOfFeatures());
        
        if(_prelearn_mode) _model->narrow_to(sub); // assume _n-dimensional model is pre-learned and needs to be narrow()ed only (NOTE: training data must not change since)
        else {_model->learn(da,sub); _model->denarrow();} // re-learn from scratch, training data has changed (i.e., after switch to next data split)
 
        // prepare feature subset index buffering
        DIMTYPE f;
        _subfeatures=0;
        for(DIMTYPE fsel : sub->selected_features()) {assert(_subfeatures<get_n()); _index[_subfeatures++]=fsel;}
        
        _model->compute_theta(); //thetas to be used in classify() and test()
        
#ifdef DEBUG
        {std::ostringstream sos; sos << "index: "; for(f=0;f<_subfeatures;f++) sos << _index[f] << " "; syncout::print(std::cout,sos);}
#endif
        assert(_subfeatures==sub->get_d_raw());
        return true;
}
 
template<class RETURNTYPE, typename DATATYPE, typename REALTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>
bool Classifier_Multinomial_NaiveBayes<RETURNTYPE,DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR>::classify(DIMTYPE &cls, const PPattern &pattern)
{
        // NOTE: train() must have been called - it learns (or narrows) the model, fills the _index
        //       map and computes the model's theta buffer this method reads. \p pattern must point
        //       at a full, raw feature vector: it is indexed through _index, not pre-narrowed.
        notify("Classifier_Multinomial_NaiveBayes::classify().");
        assert(_model);
        if(_prelearn_mode) assert(get_n()>0);
        assert(_model->get_classes()>1);
        assert(_classes==_model->get_classes());
        assert(_index);
        assert(_subfeatures==_model->get_d());
        if(!_prelearn_mode) assert(_subfeatures==_model->get_n());
        assert(_subfeatures>0);
 
        DIMTYPE f;
        DIMTYPE c_cand;
        REALTYPE res;
        DIMTYPE wTH;
        // a non-const pointer into the model's theta buffer, only read from here; taking it is safe
        // because this object owns _model exclusively and nothing reallocates the buffer during the
        // loop below (the class-major stride is _subfeatures, the model's current get_d())
        REALTYPE *theta=&(_model->get_theta()[0]);
        REALTYPE *theta_tmp;
 
        if(_subfeatures==1) { 
                // NOTE: single features are unusable for classification because theta=1 in each class,
                //       thus log(theta)=0 and consequently Pcd[] does not depend on feature frequency
                cls=0; // consider all single features equally unusable
                return false;
        } else {
                // compute P(c|pattern) for each class
                wTH=0;
                for(c_cand=0;c_cand<_classes;c_cand++)
                {
                        theta_tmp=&theta[wTH];
                        res=0.0;
                        for(f=0;f<_subfeatures;f++)
                        {
                                res+=(REALTYPE)pattern[_index[f]] * log(theta_tmp[f]);
                        }
                        _Pcd[c_cand]=log(_model->get_Pc(c_cand))+res;
                        wTH+=_subfeatures;
                }
                // find maximum P[c|pattern]
                res=_Pcd[0]; cls=0;
                for(c_cand=1;c_cand<_classes;c_cand++) if(_Pcd[c_cand]>res) {res=_Pcd[c_cand]; cls=c_cand;}
        }
        return true;
}
 
template<class RETURNTYPE, typename DATATYPE, typename REALTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>
bool Classifier_Multinomial_NaiveBayes<RETURNTYPE,DATATYPE,REALTYPE,IDXTYPE,DIMTYPE,SUBSET,DATAACCESSOR>::test(RETURNTYPE &result, const PDataAccessor da)
{
        // NOTE: train() must have been called for the subset being tested; the accessor must be
        //       positioned on the split whose TEST part is to be classified
        notify("Classifier_Multinomial_NaiveBayes::test().");
        assert(_model);
        if(_prelearn_mode) assert(get_n()>0);
        assert(_model->get_classes()>1);
        assert(_classes==_model->get_classes());
        assert(da);
        if(_prelearn_mode) assert(da->getNoOfFeatures()==get_n());
        assert(da->getNoOfClasses()>0);
        assert(da->getNoOfClasses()==_classes);
        assert(_index);
        assert(_subfeatures==_model->get_d());
        if(!_prelearn_mode) assert(_subfeatures==_model->get_n());
        assert(_subfeatures>0);
                
        typename DATAACCESSOR::PPattern p;
        IDXTYPE s,i;
        IDXTYPE count, correct;
        DIMTYPE _features=da->getNoOfFeatures();
        DIMTYPE clstmp;
 
        if(_subfeatures==1) { 
                // NOTE: single features are unusable for classification because theta=1 in each class,
                //       thus log(theta)=0 and consequently Pcd[] does not depend on feature frequency
                correct=0; count=1; // consider all single features equally unusable
        } else {
 
        
                count=0;
                correct=0;
                for(DIMTYPE c_test=0;c_test<_classes;c_test++)
                {
                        da->setClass(c_test);
                        for(const auto &blk : da->blocks(TEST)) for(i=0,p=blk.data,s=blk.size;i<s;i++)
                        {
                                if(!classify(clstmp,&p[(std::size_t)i*_features])) return false;
                                if(clstmp==c_test) correct++;
                                count++;
                        }
                }
        }
        assert(count>0);
        result=(RETURNTYPE)correct/(RETURNTYPE)count;
#ifdef DEBUG
        {
                std::ostringstream sos; sos << " result=" << result << std::endl;
                syncout::print(std::cout,sos);
        }
#endif  
        return true;
}
 
} // namespace
#endif // FSTCLASSIFIERMULTINOMNAIVEBAYES_H ///:~