/**
* Copyright (c) 2009/09-2012/08, Regents of the University of Colorado
* 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.
*/
/**
* Copyright 2012/09-2013/04, 2013/11-Present, University of Massachusetts Amherst
* Copyright 2013/05-2013/10, IPSoft Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package com.clearnlp.classification.algorithm;
import java.util.Collections;
import java.util.List;
import java.util.Random;
import com.clearnlp.classification.instance.IntInstance;
import com.clearnlp.classification.model.StringModelAD;
import com.clearnlp.classification.prediction.IntPrediction;
import com.clearnlp.classification.vector.SparseFeatureVector;
import com.clearnlp.util.UTArray;
/**
* Abstract algorithm.
* @since 1.3.2
* @author Jinho D. Choi ({@code jdchoi77@gmail.com})
*/
public class LiblinearHingeLoss extends AbstractAlgorithm
{
private final int MAX_ITER = 1000;
private double d_cost;
private double d_eps;
public LiblinearHingeLoss(double cost, double eps)
{
super(LEARN_BATCH);
init(cost, eps);
}
private void init(double cost, double eps)
{
d_cost = cost;
d_eps = eps;
}
@Override
public void train(StringModelAD model)
{
final int N = model.getInstanceSize();
double[] alpha = new double[N];
double[] qd = getQD(model);
double Gmax_old = Double.POSITIVE_INFINITY;
double Gmin_old = Double.NEGATIVE_INFINITY;
double G, d, violation, Gmax_new, Gmin_new;
int active_size = N, iter, i, s;
int [] index = UTArray.range(N);
Random rand = new Random(5);
List<IntPrediction> ps;
IntInstance instance;
IntPrediction max;
int y;
for (iter=0; iter<MAX_ITER; iter++)
{
UTArray.shuffle(rand, index, active_size);
Gmax_new = Double.NEGATIVE_INFINITY;
Gmin_new = Double.POSITIVE_INFINITY;
for (s=0; s<active_size; s++)
{
i = index[s];
instance = model.getInstance(i);
y = instance.getLabel();
ps = model.getIntPredictions(instance.getFeatureVector());
max = Collections.max(ps);
G = max.score;
if (G < 0) G = 0;
else if (max.label != y) G *= -1d;
G -= 1d;
if (alpha[i] == 0)
{
if (G > Gmax_old)
{
active_size--;
UTArray.swap(index, s, active_size);
s--;
continue;
}
violation = Math.min(G, 0);
}
else if (alpha[i] == d_cost)
{
if (G < Gmin_old)
{
active_size--;
UTArray.swap(index, s, active_size);
s--;
continue;
}
violation = Math.max(G, 0);
}
else
{
violation = G;
}
Gmax_new = Math.max(Gmax_new, violation);
Gmin_new = Math.min(Gmin_new, violation);
if (Math.abs(violation) > 1.0e-12)
{
d = alpha[i];
alpha[i] = Math.min(Math.max(d - G / qd[i], 0d), d_cost);
d = alpha[i] - d;
if (d != 0) updateWeights(model, instance, y, max.label, d);
}
}
if (Gmax_new - Gmin_new <= d_eps)
{
if (active_size == N)
break;
else
{
active_size = N;
Gmax_old = Double.POSITIVE_INFINITY;
Gmin_old = Double.NEGATIVE_INFINITY;
continue;
}
}
Gmax_old = Gmax_new;
Gmin_old = Gmin_new;
if (Gmax_old <= 0) Gmax_old = Double.POSITIVE_INFINITY;
if (Gmin_old >= 0) Gmin_old = Double.NEGATIVE_INFINITY;
System.out.print(".");
} System.out.print("\n");
}
protected double[] getQD(StringModelAD model)
{
int i, size = model.getInstanceSize();
double[] qd = new double[size];
IntInstance instance;
for (i=0; i<size; i++)
{
instance = model.getInstance(i);
qd[i] = instance.getFeatureVector().getSumOfSquaredWeights();
}
return qd;
}
private void updateWeights(StringModelAD model, IntInstance instance, int yp, int yn, double d)
{
SparseFeatureVector x = instance.getFeatureVector();
int i, xi, len = x.size();
float vi;
for (i=0; i<len; i++)
{
xi = x.getIndex(i);
vi = (float)(x.getWeight(i) * d);
model.updateWeight(yp, xi, vi);
model.updateWeight(yn, xi, -vi);
}
}
}