NeuralRectifiedLinearLayer.cpp

Go to the documentation of this file.

/*
 * Copyright (c) 2014, Shogun Toolbox Foundation
 * 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.
 *
 * 3. Neither the name of the copyright holder nor the names of its
 * contributors may be used to endorse or promote products derived from this
 * software without specific prior written permission.

 * 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 HOLDER 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.
 *
 * Written (W) 2014 Khaled Nasr
 */

#include <shogun/neuralnets/NeuralRectifiedLinearLayer.h>
#include <shogun/mathematics/Math.h>
#include <shogun/lib/SGVector.h>

using namespace shogun;

CNeuralRectifiedLinearLayer::CNeuralRectifiedLinearLayer() : CNeuralLinearLayer()
{
}

CNeuralRectifiedLinearLayer::CNeuralRectifiedLinearLayer(int32_t num_neurons):
CNeuralLinearLayer(num_neurons)
{
}

void CNeuralRectifiedLinearLayer::compute_activations(
        SGVector<float64_t> parameters,
        CDynamicObjectArray* layers)
{
    CNeuralLinearLayer::compute_activations(parameters, layers);

    int32_t len = m_num_neurons*m_batch_size;
    for (int32_t i=0; i<len; i++)
    {
        m_activations[i] = CMath::max<float64_t>(0, m_activations[i]);
    }
}

float64_t CNeuralRectifiedLinearLayer::compute_contraction_term(
    SGVector< float64_t > parameters)
{
    int32_t num_inputs = SGVector<int32_t>::sum(m_input_sizes.vector, m_input_sizes.vlen);

    SGMatrix<float64_t> W(parameters.vector+m_num_neurons,
        m_num_neurons, num_inputs, false);

    float64_t contraction_term = 0;
    for (int32_t i=0; i<m_num_neurons; i++)
    {
        float64_t sum_j = 0;
        for (int32_t j=0; j<num_inputs; j++)
            sum_j += W(i,j)*W(i,j);

        for (int32_t k = 0; k<m_batch_size; k++)
        {
            if (m_activations(i,k) > 0)
                contraction_term += sum_j;
        }
    }

    return (contraction_coefficient/m_batch_size) * contraction_term;
}

void CNeuralRectifiedLinearLayer::compute_contraction_term_gradients(
    SGVector< float64_t > parameters, SGVector< float64_t > gradients)
{
    int32_t num_inputs = SGVector<int32_t>::sum(m_input_sizes.vector, m_input_sizes.vlen);

    SGMatrix<float64_t> W(parameters.vector+m_num_neurons,
        m_num_neurons, num_inputs, false);
    SGMatrix<float64_t> WG(gradients.vector+m_num_neurons,
        m_num_neurons, num_inputs, false);

    for (int32_t k = 0; k<m_batch_size; k++)
    {
        for (int32_t i=0; i<m_num_neurons; i++)
        {
            if (m_activations(i,k) > 0)
            {
                for (int32_t j=0; j<num_inputs; j++)
                    WG(i,j) += 2 * (contraction_coefficient/m_batch_size) * W(i,j);
            }
        }
    }
}


void CNeuralRectifiedLinearLayer::compute_local_gradients(
        SGMatrix<float64_t> targets)
{
    if (targets.num_rows != 0)
    {
        int32_t length = m_num_neurons*m_batch_size;
        for (int32_t i=0; i<length; i++)
        {
            if (m_activations[i]==0)
                m_local_gradients[i] = 0;
            else
                m_local_gradients[i] = (m_activations[i]-targets[i])/m_batch_size;
        }
    }
    else
    {
        int32_t len = m_num_neurons*m_batch_size;
        for (int32_t i=0; i< len; i++)
        {
            if (m_activations[i]==0)
                m_local_gradients[i] = 0;
            else
                m_local_gradients[i] = m_activation_gradients[i];
        }
    }
}