training_rate_algorithm.h

/****************************************************************************************************************/
/*                                                                                                              */
/*   OpenNN: Open Neural Networks Library                                                                       */
/*   www.artelnics.com/opennn                                                                                   */
/*                                                                                                              */
/*   T R A I N I N G   R A T E   A L G O R I T H M   C L A S S   H E A D E R                                    */
/*                                                                                                              */
/*   Roberto Lopez                                                                                              */
/*   Artelnics - Making intelligent use of data                                                                 */
/*   [email protected]                                                                                 */
/*                                                                                                              */
/****************************************************************************************************************/

#ifndef __TRAININGRATEALGORITHM_H__
#define __TRAININGRATEALGORITHM_H__

// System includes

#include <iostream>
#include <fstream>
#include <algorithm>
#include <functional>
#include <limits>
#include <cmath>
#include <ctime>

// OpenNN includes

#include "neural_network.h"
#include "performance_functional.h"

// TinyXml includes

#include "../tinyxml2/tinyxml2.h"

namespace OpenNN
{


class TrainingRateAlgorithm
{

public:

   // ENUMERATIONS


   enum TrainingRateMethod{Fixed, GoldenSection, BrentMethod};

   // DEFAULT CONSTRUCTOR

   explicit TrainingRateAlgorithm(void);

   // GENERAL CONSTRUCTOR

   explicit TrainingRateAlgorithm(PerformanceFunctional*);

   // XML CONSTRUCTOR

   explicit TrainingRateAlgorithm(const tinyxml2::XMLDocument&);

   // DESTRUCTOR

   virtual ~TrainingRateAlgorithm(void);


   struct Triplet
   {

       Triplet(void)
       {
           A.set(2, 0.0);
           U.set(2, 0.0);
           B.set(2, 0.0);
       }


       virtual ~Triplet(void)
       {
       }


       inline bool operator == (const Triplet& other_triplet) const
       {
          if(A == other_triplet.A
          && U == other_triplet.U
          && B == other_triplet.B)
          {
             return(true);
          }
          else
          {
             return(false);
          }
       }


       inline bool has_length_zero(void) const
       {
           if(A[0] == B[0])
           {
              return(true);
           }
           else
           {
              return(false);
           }
       }


       inline std::string to_string(void) const
       {
           std::ostringstream buffer;

           buffer << "A = (" << A[0] << "," << A[1] << ")\n"
                  << "U = (" << U[0] << "," << U[1] << ")\n"
                  << "B = (" << B[0] << "," << B[1] << ")" << std::endl;

           return(buffer.str());
       }


       inline void print(void) const
       {
           std::cout << to_string();
       }


       inline void check(void) const
       {
           std::ostringstream buffer;

           if(A[0] > U[0] || U[0] > B[0])
           {
              buffer << "OpenNN Exception: TrainingRateAlgorithm class.\n"
                     << "void check(void) const method.\n"
                     << "Uncorrect triplet:\n"
                     << to_string();

              throw std::logic_error(buffer.str());
           }

           if(A[1] < U[1] || U[1] > B[1])
           {
              buffer << "OpenNN Exception: TrainingRateAlgorithm class.\n"
                     << "void check(void) const method.\n"
                     << "Triplet does not satisfy minimum condition:\n"
                     << to_string();

              throw std::logic_error(buffer.str());
           }
       }


       Vector<double> A;


       Vector<double> U;


       Vector<double> B;
   };


   // METHODS

   // Get methods

   PerformanceFunctional* get_performance_functional_pointer(void) const;

   bool has_performance_functional(void) const;

   // Training operators

   const TrainingRateMethod& get_training_rate_method(void) const;
   std::string write_training_rate_method(void) const;

   // Training parameters

   const double& get_bracketing_factor(void) const;   
   const double& get_training_rate_tolerance(void) const;

   const double& get_warning_training_rate(void) const;

   const double& get_error_training_rate(void) const;
  
   // Utilities
   
   const bool& get_display(void) const;
  
   // Set methods

   void set(void);
   void set(PerformanceFunctional*);

   void set_performance_functional_pointer(PerformanceFunctional*);

   // Training operators

   void set_training_rate_method(const TrainingRateMethod&);
   void set_training_rate_method(const std::string&);

   // Training parameters

   void set_bracketing_factor(const double&);   
   void set_training_rate_tolerance(const double&);

   void set_warning_training_rate(const double&);

   void set_error_training_rate(const double&);

   // Utilities

   void set_display(const bool&);

   virtual void set_default(void);

   // Training rate method

   double calculate_golden_section_training_rate(const Triplet&) const;
   double calculate_Brent_method_training_rate(const Triplet&) const;

   Triplet calculate_bracketing_triplet(const double&, const Vector<double>&, const double&) const;

   Vector<double> calculate_fixed_directional_point(const double&, const Vector<double>&, const double&) const;
   Vector<double> calculate_golden_section_directional_point(const double&, const Vector<double>&, const double&) const;
   Vector<double> calculate_Brent_method_directional_point(const double&, const Vector<double>&, const double&) const;

   Vector<double> calculate_directional_point(const double&, const Vector<double>&, const double&) const;

   // Serialization methods

   tinyxml2::XMLDocument* to_XML(void) const;   
   void from_XML(const tinyxml2::XMLDocument&);   


protected:

   // FIELDS


   PerformanceFunctional* performance_functional_pointer;


   // TRAINING OPERATORS


   TrainingRateMethod training_rate_method;


   double bracketing_factor;


   double training_rate_tolerance;


   double warning_training_rate;


   double error_training_rate;

   // UTILITIES


   bool display;
};

}

#endif


// OpenNN: Open Neural Networks Library.
// Copyright (c) 2005-2015 Roberto Lopez.
//
// This library is free software; you can redistribute it and/or
// modify it under the terms of the GNU Lesser General Public
// License as published by the Free Software Foundation; either
// version 2.1 of the License, or any later version.
//
// This library 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
// Lesser General Public License for more details.

// You should have received a copy of the GNU Lesser General Public
// License along with this library; if not, write to the Free Software
// Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA