ltiSupervisedInstanceClassifier.h

/*
 * Copyright (C) 2001, 2002, 2003, 2004, 2005, 2006
 * Lehrstuhl fuer Technische Informatik, RWTH-Aachen, Germany
 *
 * This file is part of the LTI-Computer Vision Library (LTI-Lib)
 *
 * The LTI-Lib is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public License (LGPL)
 * as published by the Free Software Foundation; either version 2.1 of
 * the License, or (at your option) any later version.
 *
 * The LTI-Lib 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 the LTI-Lib; see the file LICENSE.  If
 * not, write to the Free Software Foundation, Inc., 59 Temple Place -
 * Suite 330, Boston, MA 02111-1307, USA.
 */


/*----------------------------------------------------------------
 * project ....: LTI Digital Image/Signal Processing Library
 * file .......: ltiSupervisedInstanceClassifier.h
 * authors ....: Peter Doerfler
 * organization: LTI, RWTH Aachen
 * creation ...: 29.08.2001
 * revisions ..: $Id: ltiSupervisedInstanceClassifier.h,v 1.3 2006/02/07 18:26:55 ltilib Exp $
 */


#ifndef _LTI_SUPERVISEDINSTANCECLASSIFIER_H_
#define _LTI_SUPERVISEDINSTANCECLASSIFIER_H_

#include "ltiClassifier.h"
#include "ltiVector.h"
#include "ltiMatrix.h"

namespace lti {

  /**
   * Abstract class, parent of all supervised instance classifiers.
   * This class defines the interface for all supervised train methods which
   * are not dependant on time.
   */
  class supervisedInstanceClassifier : public classifier {

  public:

    /**
     * default constructor
     */
    supervisedInstanceClassifier();

    /**
     * copy constructor
     * @param other the object to be copied
     */
    supervisedInstanceClassifier(const supervisedInstanceClassifier& other);

    /**
     * returns the name of this type ("supervisedInstanceClassifier")
     */
    virtual const char* getTypeName() const;

    /**
     * copy data of "other" functor.
     * @param other the functor to be copied
     * @return a reference to this functor object
     */
    supervisedInstanceClassifier& copy(const supervisedInstanceClassifier& other);
    /**
     * Alias for "copy"
     * @param other the functor to be copied
     * @return a reference to this functor object
     */
    inline supervisedInstanceClassifier& operator=(const supervisedInstanceClassifier& other) {
      return copy(other);
    }

    /**
     * returns used parameters
     */
    const parameters& getParameters() const;

    /**
     * Supervised training.
     * The vectors in the <code>input</code> matrix
     * must be trained using as "known" classes the values given in
     * <code>ids</code>.
     * @param input the matrix with input vectors (each row is a training
     *              vector)
     * @param ids the output classes ids for the input vectors.
     * @return true if successful, false otherwise.  (if false you can check
     *              the error message with getStatusString())
     */
    virtual bool train(const dmatrix& input,
                       const ivector& ids) = 0;

    /**
     * Classification.
     * Classifies the feature and returns the outputVector with
     * the classification result.
     * @param feature the %vector to be classified
     * @param result the result of the classification
     * @return false if an error occurred during classification else true
     */
    virtual bool
      classify(const dvector& feature, outputVector& result) const =0;


  protected:

    /**
     * Sets the outputTemplate probability distributions according to
     * the classification of the given data. The distributions are
     * built by the follwing rule: <p>
     * <ol>
     * <li>Classify next data vector</li>
     * <li>For the position in the output with the highest probability
     *     increase the count for the actual id by one.</li>
     * <li>While there is more data go back to 1</li>
     * <li>For each position: divide each count by total number of counts</li>
     * </ol>
     * This results in a distribution over the ids that caused highest
     * probability for each position of the output.
     * @param outSize size of the outputTemplate
     * @param data train of validation data
     * @param ids ids of the data-vectors
     * @returns false upon error
     */
    bool makeOutputTemplate(const int& outSize,
                            const dmatrix& data,
                            const ivector& ids);
  };
}

#endif