File: structural_svm_sequence_labeling_problem.h

package info (click to toggle)
mldemos 0.5.1-3
  • links: PTS, VCS
  • area: main
  • in suites: jessie, jessie-kfreebsd
  • size: 32,224 kB
  • ctags: 46,525
  • sloc: cpp: 306,887; ansic: 167,718; ml: 126; sh: 109; makefile: 2
file content (281 lines) | stat: -rw-r--r-- 9,914 bytes parent folder | download | duplicates (4)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
// Copyright (C) 2011  Davis E. King (davis@dlib.net)
// License: Boost Software License   See LICENSE.txt for the full license.
#ifndef DLIB_STRUCTURAL_SVM_SEQUENCE_LaBELING_PROBLEM_H__
#define DLIB_STRUCTURAL_SVM_SEQUENCE_LaBELING_PROBLEM_H__


#include "structural_svm_sequence_labeling_problem_abstract.h"
#include "../matrix.h"
#include "sequence_labeler.h"
#include <vector>
#include "structural_svm_problem_threaded.h"

// ----------------------------------------------------------------------------------------

namespace dlib
{

    namespace fe_helpers
    {

    // ----------------------------------------------------------------------------------------

        struct get_feats_functor 
        {
            get_feats_functor(std::vector<std::pair<unsigned long, double> >& feats_) : feats(feats_) {}

            inline void operator() (
                unsigned long feat_index,
                double feat_value
            )
            {
                feats.push_back(std::make_pair(feat_index, feat_value));
            }

            inline void operator() (
                unsigned long feat_index
            )
            {
                feats.push_back(std::make_pair(feat_index, 1));
            }

            std::vector<std::pair<unsigned long, double> >& feats;
        };

    // ----------------------------------------------------------------------------------------

        template <typename feature_extractor, typename sequence_type, typename EXP2> 
        void get_feature_vector(
            std::vector<std::pair<unsigned long, double> >& feats,
            const feature_extractor& fe,
            const sequence_type& sequence,
            const matrix_exp<EXP2>& candidate_labeling,
            unsigned long position
        )
        {
            get_feats_functor funct(feats);
            fe.get_features(funct, sequence,candidate_labeling, position);
        }

    }

// ----------------------------------------------------------------------------------------

    template <
        typename feature_extractor
        >
    class structural_svm_sequence_labeling_problem : noncopyable,
        public structural_svm_problem_threaded<matrix<double,0,1>, std::vector<std::pair<unsigned long,double> > >
    {
    public:
        typedef matrix<double,0,1> matrix_type;
        typedef std::vector<std::pair<unsigned long, double> > feature_vector_type;

        typedef typename feature_extractor::sequence_type sequence_type;

        structural_svm_sequence_labeling_problem(
            const std::vector<sequence_type>& samples_,
            const std::vector<std::vector<unsigned long> >& labels_,
            const feature_extractor& fe_,
            unsigned long num_threads = 2
        ) :
            structural_svm_problem_threaded<matrix_type,feature_vector_type>(num_threads),
            samples(samples_),
            labels(labels_),
            fe(fe_)
        {
            // make sure requires clause is not broken
            DLIB_ASSERT(is_sequence_labeling_problem(samples,labels) == true &&
                        contains_invalid_labeling(fe, samples, labels) == false,
                        "\t structural_svm_sequence_labeling_problem::structural_svm_sequence_labeling_problem()"
                        << "\n\t invalid inputs were given to this function"
                        << "\n\t samples.size(): " << samples.size() 
                        << "\n\t is_sequence_labeling_problem(samples,labels): " << is_sequence_labeling_problem(samples,labels)
                        << "\n\t contains_invalid_labeling(fe,samples,labels): " << contains_invalid_labeling(fe,samples,labels)
                        << "\n\t this: " << this
                        );

#ifdef ENABLE_ASSERTS
            for (unsigned long i = 0; i < labels.size(); ++i)
            {
                for (unsigned long j = 0; j < labels[i].size(); ++j)
                {
                    // make sure requires clause is not broken
                    DLIB_ASSERT(labels[i][j] < fe.num_labels(),
                                "\t structural_svm_sequence_labeling_problem::structural_svm_sequence_labeling_problem()"
                                << "\n\t The given labels in labels are invalid."
                                << "\n\t labels[i][j]: " << labels[i][j] 
                                << "\n\t fe.num_labels(): " << fe.num_labels()
                                << "\n\t i: " << i 
                                << "\n\t j: " << j 
                                << "\n\t this: " << this
                                );
                }
            }
#endif

            loss_values.assign(num_labels(), 1);

        }

        unsigned long num_labels (
        ) const { return fe.num_labels(); }

        double get_loss (
            unsigned long label
        ) const 
        { 
            // make sure requires clause is not broken
            DLIB_ASSERT(label < num_labels(),
                        "\t void structural_svm_sequence_labeling_problem::get_loss()"
                        << "\n\t invalid inputs were given to this function"
                        << "\n\t label:        " << label 
                        << "\n\t num_labels(): " << num_labels() 
                        << "\n\t this:         " << this
                        );

            return loss_values[label]; 
        }

        void set_loss (
            unsigned long label,
            double value
        )  
        { 
            // make sure requires clause is not broken
            DLIB_ASSERT(label < num_labels() && value >= 0,
                        "\t void structural_svm_sequence_labeling_problem::set_loss()"
                        << "\n\t invalid inputs were given to this function"
                        << "\n\t label:        " << label 
                        << "\n\t num_labels(): " << num_labels() 
                        << "\n\t value:        " << value 
                        << "\n\t this:         " << this
                        );

            loss_values[label] = value;
        }

    private:
        virtual long get_num_dimensions (
        ) const 
        {
            return fe.num_features();
        }

        virtual long get_num_samples (
        ) const 
        {
            return samples.size();
        }

        void get_joint_feature_vector (
            const sequence_type& sample, 
            const std::vector<unsigned long>& label,
            feature_vector_type& psi
        ) const 
        {
            psi.clear();

            const int order = fe.order();

            matrix<unsigned long,0,1> candidate_labeling; 
            for (unsigned long i = 0; i < sample.size(); ++i)
            {
                candidate_labeling = rowm(vector_to_matrix(label), range(i, std::max((int)i-order,0)));

                fe_helpers::get_feature_vector(psi,fe,sample,candidate_labeling, i);
            }
        }

        virtual void get_truth_joint_feature_vector (
            long idx,
            feature_vector_type& psi 
        ) const 
        {
            get_joint_feature_vector(samples[idx], labels[idx], psi);
        }

        class map_prob
        {
        public:
            unsigned long order() const { return fe.order(); }
            unsigned long num_states() const { return fe.num_labels(); }

            map_prob(
                const sequence_type& sequence_,
                const std::vector<unsigned long>& label_,
                const feature_extractor& fe_,
                const matrix<double,0,1>& weights_,
                const std::vector<double>& loss_values_
            ) :
                sequence(sequence_),
                label(label_),
                fe(fe_),
                weights(weights_),
                loss_values(loss_values_)
            {
            }

            unsigned long number_of_nodes(
            ) const
            {
                return sequence.size();
            }

            template <
                typename EXP 
                >
            double factor_value (
                unsigned long node_id,
                const matrix_exp<EXP>& node_states
            ) const
            {
                if (dlib::impl::call_reject_labeling_if_exists(fe, sequence,  node_states, node_id))
                    return -std::numeric_limits<double>::infinity();

                double loss = 0;
                if (node_states(0) != label[node_id])
                    loss = loss_values[label[node_id]];

                return fe_helpers::dot(weights, fe, sequence, node_states, node_id) + loss;
            }

            const sequence_type& sequence;
            const std::vector<unsigned long>& label;
            const feature_extractor& fe;
            const matrix<double,0,1>& weights;
            const std::vector<double>& loss_values;
        };

        virtual void separation_oracle (
            const long idx,
            const matrix_type& current_solution,
            scalar_type& loss,
            feature_vector_type& psi
        ) const
        {
            std::vector<unsigned long> y;
            find_max_factor_graph_viterbi(map_prob(samples[idx],labels[idx],fe,current_solution,loss_values), y);

            loss = 0;
            for (unsigned long i = 0; i < y.size(); ++i)
            {
                if (y[i] != labels[idx][i])
                    loss += loss_values[labels[idx][i]];
            }

            get_joint_feature_vector(samples[idx], y, psi);
        }

        const std::vector<sequence_type>& samples;
        const std::vector<std::vector<unsigned long> >& labels;
        const feature_extractor& fe;
        std::vector<double> loss_values;
    };

// ----------------------------------------------------------------------------------------

}

#endif // DLIB_STRUCTURAL_SVM_SEQUENCE_LaBELING_PROBLEM_H__