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
|
/**
* Define some weak classifiers inherited from abstract Classifier class.
*/
#ifndef CLASSIFIER_H
#define CLASSIFIER_H
#include <math.h>
#include "data.h"
#include "random.h"
template<class C, class dataT, class labelT>
class Classifier
{
public:
// randomly get a classifier
virtual C RandomClassifier(Random& random) = 0;
// calculate feature's response for this classifier
virtual double FeatureResponse(const DataSet& data, index_t index) const = 0;
// calculate boolean response for this classifier and threshold
bool Response(const DataSet& data, index_t index)
{
return FeatureResponse(data, index) < threshold_? true:false;
}
virtual void Read(std::istream& is) = 0;
virtual void Write(std::ostream& os) = 0;
double threshold_;
};
template<class dataT, class labelT>
class AxisAlignedClassifier :
public Classifier<AxisAlignedClassifier<dataT, labelT>, dataT, labelT>
{
public:
typedef Classifier<AxisAlignedClassifier<dataT, labelT>, dataT, labelT> ClassifierT;
typedef AxisAlignedClassifier<dataT, labelT> AxisAlignedClassifierT;
typedef MLData<dataT, labelT> MLDataT;
AxisAlignedClassifier(): featureDim_(-1), axis_(-1) {}
AxisAlignedClassifier(int dimension): featureDim_(dimension), axis_(-1) {}
AxisAlignedClassifier(int dimension, int axis): featureDim_(dimension), axis_(axis) {}
AxisAlignedClassifierT RandomClassifier(Random &random)
{
return AxisAlignedClassifierT(featureDim_, random.RandI(0, featureDim_));
}
double FeatureResponse(const DataSet& data, index_t index) const
{
return ((const MLDataT&)data).data[index][axis_];
}
void Print(int level)
{
std::cout << "- Classifier: axis"
<< " dim = " << featureDim_
<< " axis = " << axis_
<< " threshold = " << ClassifierT::threshold_;
if ((level/100 - (level/1000)*10) == 2)
{
std::cout << " [Addr: " << this << "]";
}
std::cout << std::endl;
}
virtual void Read(std::istream& is)
{
readBasicType(is, featureDim_);
readBasicType(is, axis_);
readBasicType(is, ClassifierT::threshold_);
}
virtual void Write(std::ostream& os)
{
writeBasicType(os, featureDim_);
writeBasicType(os, axis_);
writeBasicType(os, ClassifierT::threshold_);
}
int featureDim_;
int axis_;
};
template<class dataT, class labelT>
class LinearClassifier
: public Classifier<LinearClassifier<dataT, labelT>, dataT, labelT>
{
public:
typedef Classifier<LinearClassifier<dataT, labelT>, dataT, labelT> ClassifierT;
typedef LinearClassifier<dataT, labelT> LinearClassifierT;
typedef MLData<dataT, labelT> MLDataT;
LinearClassifier(): featureDim_(-1) {}
LinearClassifier(int dimension): featureDim_(dimension)
{
unitVector_.resize(dimension);
}
LinearClassifier(int dimension, std::vector<dataT>& unitVector)
: featureDim_(dimension), unitVector_(unitVector) {}
LinearClassifierT RandomClassifier(Random &random)
{
double length = 0;
for (int i = 0; i < featureDim_; ++i)
{
unitVector_[i] = 2.0 * random.RandD() - 1.0;
length += unitVector_[i] * unitVector_[i];
}
length = sqrt(length);
for (int i = 0; i < featureDim_; ++i)
{
unitVector_[i] = unitVector_[i] / length;
}
return LinearClassifierT(featureDim_, unitVector_);
}
double FeatureResponse(const DataSet &data, index_t index) const
{
double dotProduct = 0;
for (int i = 0; i < featureDim_; ++i)
{
dotProduct += ((const MLDataT&)data).data[index][i] * unitVector_[i];
}
return dotProduct;
}
void Print(int level)
{
std::cout << "- Classifier: linear"
<< " dim = " << featureDim_
<< " unit vector = [";
for (int i = 0; i < featureDim_; ++i)
{
std::cout << unitVector_[i];
if (i != (featureDim_-1))
{
std::cout << " ";
}
else
{
std::cout << "]";
}
}
std::cout << " threshold = " << ClassifierT::threshold_;
if ((level/100 - (level/1000)*10) == 2)
{
std::cout << " [Addr: " << this << "]";
}
std::cout << std::endl;
}
virtual void Read(std::istream& is)
{
readBasicType(is, featureDim_);
unitVector_.resize(featureDim_);
for (int i = 0; i < featureDim_; ++i)
{
readBasicType(is, unitVector_[i]);
}
readBasicType(is, ClassifierT::threshold_);
}
virtual void Write(std::ostream& os)
{
writeBasicType(os, featureDim_);
for (int i = 0; i < featureDim_; ++i)
{
writeBasicType(os, unitVector_[i]);
}
writeBasicType(os, ClassifierT::threshold_);
}
int featureDim_;
std::vector<dataT> unitVector_;
};
#endif // CLASSIFIER_H
|
