SoftMaxLikelihood.h

Go to the documentation of this file.

/*
 * Copyright (c) The Shogun Machine Learning Toolbox
 * Written (w) 2014 Wu Lin
 * Written (w) 2014 Parijat Mazumdar
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 * 1. Redistributions of source code must retain the above copyright notice, this
 * list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright notice,
 * this list of conditions and the following disclaimer in the documentation
 * and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 *
 * The views and conclusions contained in the software and documentation are those
 * of the authors and should not be interpreted as representing official policies,
 * either expressed or implied, of the Shogun Development Team.
 *
 * Code adapted from
 * https://gist.github.com/yorkerlin/8a36e8f9b298aa0246a4
 * and
 * GPstuff - Gaussian process models for Bayesian analysis
 * http://becs.aalto.fi/en/research/bayes/gpstuff/
 *
 * The reference pseudo code is the algorithm 3.4 of the GPML textbook
 *
 */

#ifndef _SOFTMAXLIKELIHOOD_H_
#define _SOFTMAXLIKELIHOOD_H_

#include <shogun/lib/config.h>

#ifdef HAVE_EIGEN3

#include <shogun/labels/MulticlassLabels.h>
#include <shogun/machine/gp/LikelihoodModel.h>

namespace shogun
{

enum EMCSamplerType
{
    MC_Probability,
    MC_Mean,
    MC_Variance
};

class CSoftMaxLikelihood : public CLikelihoodModel
{
public:
    CSoftMaxLikelihood();

    virtual ~CSoftMaxLikelihood();

    virtual const char* get_name() const { return "SoftMaxLikelihood"; }

    virtual SGVector<float64_t> get_predictive_means(SGVector<float64_t> mu,
            SGVector<float64_t> s2, const CLabels* lab=NULL) const;

    virtual SGVector<float64_t> get_predictive_variances(SGVector<float64_t> mu,
            SGVector<float64_t> s2, const CLabels* lab=NULL) const;

    virtual SGVector<float64_t> get_predictive_log_probabilities(SGVector<float64_t> mu,
        SGVector<float64_t> s2, const CLabels *lab=NULL);

    virtual SGVector<float64_t> get_log_probability_f(const CLabels* lab,
            SGVector<float64_t> func) const;

    virtual SGVector<float64_t> get_log_probability_derivative_f(
            const CLabels* lab, SGVector<float64_t> func, index_t i) const;

    virtual SGVector<float64_t> get_log_zeroth_moments(SGVector<float64_t> mu,
            SGVector<float64_t> s2, const CLabels* lab) const
    {
        SG_ERROR("Not Implemented\n");
        return SGVector<float64_t>();
    }

    virtual float64_t get_first_moment(SGVector<float64_t> mu,
            SGVector<float64_t> s2, const CLabels* lab, index_t i) const
    {
        SG_ERROR("Not Implemented\n");
        return -1.0;
    }

    virtual float64_t get_second_moment(SGVector<float64_t> mu,
            SGVector<float64_t> s2, const CLabels* lab, index_t i) const
    {
        SG_ERROR("Not Implemented\n");
        return -1.0;
    }

    virtual bool supports_multiclass() const { return true; }

    virtual void set_num_samples(index_t num_samples);

private:
    void init();
    index_t m_num_samples;

    SGVector<float64_t> predictive_helper(SGVector<float64_t> mu,
    SGVector<float64_t> s2, const CLabels *lab, EMCSamplerType option) const;

    SGVector<float64_t> mc_sampler(index_t num_samples, SGVector<float64_t> mean,
        SGMatrix<float64_t> Sigma, SGVector<float64_t> y) const;


    SGVector<float64_t> get_log_probability_derivative1_f(const CLabels* lab, SGMatrix<float64_t> func) const;

    SGVector<float64_t> get_log_probability_derivative2_f(SGMatrix<float64_t> func) const;

    SGVector<float64_t> get_log_probability_derivative3_f(SGMatrix<float64_t> func) const;
};
}
#endif /* HAVE_EIGEN3 */
#endif /* _SOFTMAXLIKELIHOOD_H_ */