forked from viccwq/ANN
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathSampleNN.cpp
More file actions
67 lines (60 loc) · 1.46 KB
/
SampleNN.cpp
File metadata and controls
67 lines (60 loc) · 1.46 KB
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
#include "SampleNN.h"
//http://www.cnblogs.com/ronny/p/ann_01.html
//http://www.cnblogs.com/ronny/p/ann_02.html
SamleNN::SamleNN(void)
: actual_output(0)
{
getSamplesData();
initialWgt();
}
SamleNN::~SamleNN(void)
{
inputData.clear();
weight.clear();
}
//样本获取：
void SamleNN::getSamplesData(void)
{
const int iterations = 15000; // 15000个样本
for (int i = 0; i < iterations; i++)
{
int index = i % 4;
vector<double> dvect(4, 0);
dvect[index] = 1;
for (size_t i = 0; i != dvect.size(); i++)
{
dvect[i] += (5e-3*rand() / RAND_MAX - 2.5e-3);
}
inputData.push_back(dvect);
}
}
//用[0,0.05]之间的随机值初始化权重。
void SamleNN::initialWgt(void)
{
// 4个连结和一个偏置w0
for (int i = 0; i != 5; i++)
{
weight.push_back(0.05*rand()/RAND_MAX);
}
}
//向前计算
void SamleNN::cmtForward(const vector<double>& inVect)
{
double dsum = weight[4];//先把偏置加上
for (size_t i = 0; i != inVect.size(); i++)
{
dsum += (inVect[i] * weight[i]);
}
actual_output = 1 / (1 + exp(-1*dsum));
}
//更新权重
void SamleNN::updataWgt(const vector<double>& inVect, const double true_output)
{
double learnRate = 0.05; // 权重更新参数
for (size_t i = 0; i != weight.size() - 1; i++)
{
weight[i] += (learnRate*(true_output - actual_output)*actual_output*(1 - actual_output)*inVect[i]);
}
// w0单独计算
weight[4] += (learnRate*(true_output - actual_output)*actual_output*(1 - actual_output)*1);
}